def create_objects(cls, securities):
ranks = utils.get_pumps_rank()
quotient_metrics = utils.get_quotient_metrics()
ranks['quotient'] = quotient_metrics['quotient']
foo = lambda x: utils.spam_counts.get(x, 0)
ranks['spams'] = map(foo, ranks['symbol'])
ranks = ranks.sort('quotient', ascending=False)
cls._pre_filtered_ranks = {
'All': {k: ranks[k]
for k in ranks.columns},
'Stocks with Spam':
dict(ranks[ranks['symbol'].isin(plugins.spam_counts.keys())].sort(
'quotient', ascending=False)),
'Stocks without Spam':
dict(ranks[~ranks['symbol'].isin(plugins.spam_counts.keys())].sort(
'quotient', ascending=False)),
}
source_stocks_rank = ColumnDataSource(
cls._pre_filtered_ranks['Stocks with Spam'])
table_stocks_rank = DataTable(source=source_stocks_rank,
width=785,
height=400,
selectable=True,
editable=True,
columns=[
TableColumn(field='symbol',
title='symbol',
editor=StringEditor()),
TableColumn(field='quotient',
title='quotient',
editor=StringEditor()),
TableColumn(field='rank',
title='rank',
editor=StringEditor()),
TableColumn(field='spams',
title='spams',
editor=StringEditor()),
])
callback = Callback(args={
'tr': table_stocks_rank,
'sr': source_stocks_rank
},
code=callbacks.source_stocks_rank)
source_stocks_rank.callback = callback
return locals()
def draw_bokeh(vecs, sentences):
""" Dimension reduction with PCA and plotting with bokeh """
X = np.array([x["feature_vector"] for x in vecs])
X_reduced = PCA(n_components=2).fit_transform(X)
source = ColumnDataSource(data=dict(
eigenv1=[x[0] for x in X_reduced],
eigenv2=[x[1] for x in X_reduced],
s=sentences,
))
hover = HoverTool(tooltips=[("index",
"$index"), ("(x,y)",
"@eigenv1, @eigenv2)"), ("s",
"@s")])
p = figure(title="Guardia nacional",
tools=[hover, ResetTool(), BoxSelectTool()])
p.xaxis.axis_label = '1st eigenv'
p.yaxis.axis_label = '2nd eigenv'
p.circle('eigenv1',
'eigenv2',
size=10,
fill_alpha=1,
nonselection_fill_alpha=0.5,
source=source)
div = Div(width=400)
layout = row(p, div)
source.callback = CustomJS(args=dict(source=source,
div=div,
sentences=sentences),
code="""
var inds = source.selected.indices
div.text = ""
for (var i = 0; i < sentences.length; i++) {
if (inds.includes(i)) {
div.text += '<p style="background-color:powderblue;">' + sentences[i] + "</p>"
}
else {
div.text += "<p>" + sentences[i] + "</p>"
}
}
""")
output_file("news.html", title="Sentence embedding example")
show(layout)
def create_objects(cls, securities):
ranks = utils.get_pumps_rank()
quotient_metrics = utils.get_quotient_metrics()
ranks['quotient'] = quotient_metrics['quotient']
foo = lambda x: utils.spam_counts.get(x, 0)
ranks['spams'] = map(foo, ranks['symbol'])
ranks = ranks.sort('quotient', ascending=False)
cls._pre_filtered_ranks = {
'All': {k: ranks[k] for k in ranks.columns},
'Stocks with Spam': dict(ranks[ranks['symbol'].
isin(plugins.spam_counts.keys())].
sort('quotient', ascending=False)),
'Stocks without Spam': dict(ranks[~ranks['symbol'].
isin(plugins.spam_counts.keys())].
sort('quotient', ascending=False)),
}
source_stocks_rank = ColumnDataSource(cls._pre_filtered_ranks['Stocks with Spam'])
table_stocks_rank = DataTable(
source=source_stocks_rank, width=785, height=400,
selectable=True, editable=True,
columns=[
TableColumn(field='symbol', title='symbol', editor=StringEditor()),
TableColumn(field='quotient', title='quotient', editor=StringEditor()),
TableColumn(field='rank', title='rank', editor=StringEditor()),
TableColumn(field='spams', title='spams', editor=StringEditor()),
])
callback = Callback(
args={'tr': table_stocks_rank, 'sr': source_stocks_rank},
code=callbacks.source_stocks_rank
)
source_stocks_rank.callback = callback
return locals()
def create_objects(cls, symbol, df, securities):
descr_box = Paragraph(text='content loading...')
btn_close_loading = Button(label='Close Loading')
dialog_loading = Dialog(
title='loading', content=vplot(descr_box), name='loading_dialog',
buttons=[btn_close_loading], visible=False)
source_data = dict(df)
main_source = ColumnDataSource(dict(df))
source = ColumnDataSource(source_data)
# TODO: REMOVE THIS COMMENTED CODE! IT'S JUST THE PREVIOUS
# VERSION USED BEFORE NEW P&D Cached results and algorithm
# get the cached results of the P&D algorithm computed with the
# "default" configuration
# intervals = utils.cached_pumps.get(symbol, pumps.to_dicts(((),(),(),(),(),())))
# intervals['bottom'] = [0] * len(intervals['start'])
# intervals['values'] = [max(df['price'])] * len(intervals['start'])
#
# intervals = pd.DataFrame(intervals)
# new version
stats = utils.get_symbols_cached_stats()[symbol]
intervals = pd.DataFrame(stats)
intervals['bottom'] = [0] * len(intervals['start'])
intervals['values'] = [max(df['price'])] * len(intervals['start'])
conv = lambda x: utils.to_seconds(pd.to_datetime(x))
intervals = intervals[
(pd.to_datetime(intervals['start']) > conv(config.date_range[0])) &
(pd.to_datetime(intervals['start']) < conv(config.date_range[1]))
]
# Create P&Ds intervals DataSource
intervals_source = ColumnDataSource(intervals)
source.tags = ['main_source']
trends = utils.load_trends_data(symbol, start_date=min(df['dt']))
trends_source = ColumnDataSource(trends)
trades = Slider(
title="trades", name='trades',
value=0, start=0, end=124, step=1
)
# Selectors
symbol = Select.create(
options=securities, value=symbol, name='symbol', title=""
)
window_selector = Select.create(
options=['---'], name='period_selector', title="Search intervals with:"
)
symbol_filter = Select.create(
options=['All', 'Stocks with Spam', 'Stocks without Spam'],
name='symbol_filter', title="Filter Symbols:",
value='Stocks with Spam'
)
callback = Callback(
args={'symbol_filter': symbol_filter,
'dialog_loading': dialog_loading},
code=callbacks.symbol_filter
)
symbol_filter.callback = callback
btn_detect_pumps = Button(label='Configure P&D Detection', name='config_pumps')
main_tab = Panel(title="Main")
tabs = Tabs()
# Create STOCKS TABLE
ranks = utils.get_pumps_rank()
# quotient_metrics = utils.get_quotient_metrics()
# ranks['quotient'] = quotient_metrics['quotient']
foo = lambda x: utils.spams_count.get(x, 0)
ranks['spams'] = map(foo, ranks['symbol'])
ranks = ranks.sort(['spams', 'vol_quotient'], ascending=False)
cls._pre_filtered_ranks = {
'All': {k: ranks[k] for k in ranks.columns},
'Stocks with Spam': dict(ranks[ranks['spams'] > 0].
sort('vol_quotient', ascending=False)),
'Stocks without Spam': dict(ranks[ranks['spams'] == 0].
sort('vol_quotient', ascending=False)),
}
source_stocks_rank = ColumnDataSource(cls._pre_filtered_ranks['All'])
table_stocks_rank = DataTable(
source=source_stocks_rank, width=560, height=450,
selectable=True, editable=True,
columns=[
TableColumn(field='symbol', title='symbol', width=130, editor=StringEditor()),
TableColumn(field='vol_quotient', title='volume ratio', editor=StringEditor(),
default_sort='descending'),
TableColumn(field='risk_score', title='risk', width=100, editor=StringEditor(),
default_sort='descending'),
TableColumn(field='spams', title='spams', width=130, editor=StringEditor(),
default_sort='descending'),
])
callback = Callback(args={'tr': table_stocks_rank, 'sr': source_stocks_rank, 'symb': symbol,
'dialog_loading': dialog_loading},
code=callbacks.source_stocks_rank)
source_stocks_rank.callback = callback
return locals()
def make_dashboard(motors, xy_train, operating_earnings, maintenance_cost,
repair_cost, run_interval):
#calculate revenue vs time dataframe
print '...generating dashboard...'
events = get_events(motors)
events['earnings'] = 0.0
events.loc[events.state == 'operating', 'earnings'] = operating_earnings
events['expenses'] = 0.0
events.loc[events.state == 'maintenance', 'expenses'] = maintenance_cost
events.loc[events.state == 'repair', 'expenses'] = repair_cost
money = events.groupby('Time').sum()[['earnings', 'expenses']]
money['revenue'] = money.earnings - money.expenses
money['cumulative_earnings'] = money.earnings.cumsum()
money['cumulative_expenses'] = money.expenses.cumsum()
money['cumulative_revenue'] = money.revenue.cumsum()
#map the (P,T) decision surface
T_min = 50
T_max = 150
P_min = 0
P_max = 100
T_axis = np.arange(T_min, T_max, 0.5)
P_axis = np.arange(P_min, P_max, 0.5)
x, y = np.meshgrid(T_axis, P_axis)
ttf = np.zeros((len(P_axis), len(T_axis)))
import copy
m = copy.deepcopy(motors[0])
for p_idx in np.arange(len(P_axis)):
for t_idx in np.arange(len(T_axis)):
m.Temp = T_axis[t_idx]
m.Pressure = P_axis[p_idx]
ttf[p_idx, t_idx] = m.predicted_time_to_fail()
#plot decision surface
from bokeh.plotting import figure, show, output_file, ColumnDataSource, vplot
from bokeh.models import HoverTool, Callback
from bokeh.io import vform
output_file('dashboard.html', title='Smart Maintenance Dashboard')
source = ColumnDataSource(data=dict(
x=xy_train.Temp,
y=xy_train.Pressure,
ttf=xy_train.time_to_fail,
size=0.6 * xy_train.time_to_fail,
))
dec_fig = figure(x_range=[T_min, T_max],
y_range=[P_min, P_max],
title='SVM Decision Surface',
x_axis_label='Temperature',
y_axis_label='Pressure',
tools='box_zoom,reset,hover,crosshair',
width=600,
plot_height=600)
dec_fig.title_text_font_size = '18pt'
dec_fig.xaxis.axis_label_text_font_size = '14pt'
dec_fig.yaxis.axis_label_text_font_size = '14pt'
dec_fig.image(image=[-ttf],
x=[T_min],
y=[P_min],
dw=[T_max - T_min],
dh=[P_max - P_min],
palette='RdYlGn8')
dec_fig.x('x',
'y',
size='size',
source=source,
fill_alpha=0.5,
fill_color='navy',
line_color='navy',
line_width=1,
line_alpha=0.5)
hover = dec_fig.select(dict(type=HoverTool))
hover.tooltips = [
("Temperature", "@x"),
("Pressure", "@y"),
("measured lifetime", "@ttf"),
]
#plot earnings vs time
source = ColumnDataSource(data=dict(
t=money.index,
earnings=money.cumulative_earnings / 1.e6,
expenses=money.cumulative_expenses / 1.e6,
revenue=money.cumulative_revenue / 1.e6,
zero=money.cumulative_revenue * 0,
))
earn_fig = figure(title='Cumulative Earnings & Expenses',
x_axis_label='Time',
y_axis_label='Earnings & Expenses (M$)',
tools='box_zoom,reset,hover,crosshair',
width=1000,
plot_height=300,
x_range=[0, 1200],
y_range=[0, 120])
earn_fig.title_text_font_size = '15pt'
earn_fig.xaxis.axis_label_text_font_size = '11pt'
earn_fig.yaxis.axis_label_text_font_size = '11pt'
earn_fig.line('t',
'earnings',
color='blue',
source=source,
line_width=5,
legend='earnings')
earn_fig.line('t',
'expenses',
color='red',
source=source,
line_width=5,
legend='expenses')
earn_fig.legend.orientation = "bottom_right"
earn_fig.patch([0, 200, 200, 0], [0, 0, 120, 120],
color='lightsalmon',
alpha=0.35,
line_width=0)
earn_fig.patch([200, 400, 400, 200], [0, 0, 120, 120],
color='gold',
alpha=0.35,
line_width=0)
earn_fig.patch([400, 1200, 1200, 400], [0, 0, 120, 120],
color='darkseagreen',
alpha=0.35,
line_width=0)
earn_fig.text([45], [101], ['run-to-fail'])
earn_fig.text([245], [101], ['scheduled'])
earn_fig.text([245], [90], ['maintenance'])
earn_fig.text([445], [101], ['predictive'])
earn_fig.text([445], [90], ['maintenance'])
hover = earn_fig.select(dict(type=HoverTool))
hover.tooltips = [
(" Time", "@t"),
(" earning (M$)", "@earnings"),
("expenses (M$)", "@expenses"),
]
#plot revenue vs time
rev_fig = figure(title='Cumulative Revenue',
x_axis_label='Time',
y_axis_label='Revenue (M$)',
tools='box_zoom,reset,hover,crosshair',
width=1000,
plot_height=300,
x_range=[0, 1200],
y_range=[-15, 10])
rev_fig.title_text_font_size = '15pt'
rev_fig.xaxis.axis_label_text_font_size = '11pt'
rev_fig.yaxis.axis_label_text_font_size = '11pt'
rev_fig.line('t',
'revenue',
color='green',
source=source,
line_width=5,
legend='revenue')
rev_fig.line('t',
'zero',
color='purple',
source=source,
line_width=3,
alpha=0.5,
line_dash=[10, 5])
rev_fig.legend.orientation = "bottom_right"
rev_fig.patch([0, 200, 200, 0], [-15, -15, 10, 10],
color='lightsalmon',
alpha=0.35,
line_width=0)
rev_fig.patch([200, 400, 400, 200], [-15, -15, 10, 10],
color='gold',
alpha=0.35,
line_width=0)
rev_fig.patch([400, 1200, 1200, 400], [-15, -15, 10, 10],
color='darkseagreen',
alpha=0.35,
line_width=0)
hover = rev_fig.select(dict(type=HoverTool))
hover.tooltips = [
(" Time", "@t"),
(" revenue (M$)", "@revenue"),
]
#plot number of motors vs time
N = events.groupby(['Time', 'state']).count().unstack()['id'].reset_index()
N.fillna(value=0, inplace=True)
N['total'] = N.maintenance + N.operating + N.repair
s1 = ColumnDataSource(data=dict(
Time=N.Time,
operating=N.operating,
maintenance=N.maintenance,
repair=N.repair,
total=N.total,
))
motor_fig = figure(title='Number of Motors',
x_axis_label='Time',
y_axis_label='Number of motors',
tools='box_zoom,reset,hover,crosshair',
width=1000,
plot_height=300,
x_range=[0, 1200],
y_range=[-10, 210])
motor_fig.title_text_font_size = '15pt'
motor_fig.xaxis.axis_label_text_font_size = '11pt'
motor_fig.yaxis.axis_label_text_font_size = '11pt'
motor_fig.line('Time',
'total',
color='blue',
source=s1,
line_width=3,
legend='total')
motor_fig.line('Time',
'operating',
color='green',
source=s1,
line_width=3,
legend='operating')
motor_fig.line('Time',
'maintenance',
color='orange',
source=s1,
line_width=3,
legend='maintenance')
motor_fig.line('Time',
'repair',
color='red',
source=s1,
line_width=3,
legend='repair')
motor_fig.legend.orientation = "top_right"
motor_fig.patch([0, 200, 200, 0], [-10, -10, 210, 210],
color='lightsalmon',
alpha=0.35,
line_width=0)
motor_fig.patch([200, 400, 400, 200], [-10, -10, 210, 210],
color='gold',
alpha=0.35,
line_width=0)
motor_fig.patch([400, 1200, 1200, 400], [-10, -10, 210, 210],
color='darkseagreen',
alpha=0.35,
line_width=0)
#display N table
from bokeh.models.widgets import DataTable, TableColumn
from bokeh.io import vform
columns = [
TableColumn(field='Time', title='Time'),
TableColumn(field='operating', title='operating'),
TableColumn(field='maintenance', title='maintenance'),
TableColumn(field='repair', title='repair'),
TableColumn(field='total', title='total'),
]
s2 = s1.clone()
N_table = DataTable(source=s2, columns=columns, width=600, height=300)
s1.callback = Callback(args=dict(s2=s2),
code="""
var inds = cb_obj.get('selected')['1d'].indices;
var d1 = cb_obj.get('data');
var d2 = s2.get('data');
d2['Time'] = []
d2['operating'] = []
d2['maintenance'] = []
d2['repair'] = []
d2['total'] = []
for (i = 0; i < inds.length; i++) {
d2['Time'].push(d1['Time'][inds[i]])
d2['operating'].push(d1['operating'][inds[i]])
d2['maintenance'].push(d1['maintenance'][inds[i]])
d2['repair'].push(d1['repair'][inds[i]])
d2['total'].push(d1['total'][inds[i]])
}
s2.trigger('change');
""")
#export plot to html and return
plot_grid = vplot(dec_fig, earn_fig, rev_fig, motor_fig, vform(N_table))
show(plot_grid, new='tab')
return money, events, N
def plot_3DModel_bokeh(filename, map_data_all_slices, map_depth_all_slices, \
color_range_all_slices, profile_data_all, boundary_data, \
style_parameter):
'''
Plot shear velocity maps and velocity profiles using bokeh
Input:
filename is the filename of the resulting html file
map_data_all_slices contains the velocity model parameters saved for map view plots
map_depth_all_slices is a list of depths
color_range_all_slices is a list of color ranges
profile_data_all constains the velocity model parameters saved for profile plots
boundary_data is a list of boundaries
style_parameter contains plotting parameters
Output:
None
'''
xlabel_fontsize = style_parameter['xlabel_fontsize']
#
colorbar_data_all_left = []
colorbar_data_all_right = []
map_view_ndepth = style_parameter['map_view_ndepth']
ncolor = len(palette)
colorbar_top = [0.1 for i in range(ncolor)]
colorbar_bottom = [0 for i in range(ncolor)]
map_data_all_slices_depth = []
for idepth in range(map_view_ndepth):
color_min = color_range_all_slices[idepth][0]
color_max = color_range_all_slices[idepth][1]
color_step = (color_max - color_min)*1./ncolor
colorbar_left = np.linspace(color_min,color_max-color_step,ncolor)
colorbar_right = np.linspace(color_min+color_step,color_max,ncolor)
colorbar_data_all_left.append(colorbar_left)
colorbar_data_all_right.append(colorbar_right)
map_depth = map_depth_all_slices[idepth]
map_data_all_slices_depth.append('Depth: {0:8.0f} km'.format(map_depth))
#
palette_r = palette[::-1]
# data for the colorbar
colorbar_data_one_slice = {}
colorbar_data_one_slice['colorbar_left'] = colorbar_data_all_left[style_parameter['map_view_default_index']]
colorbar_data_one_slice['colorbar_right'] = colorbar_data_all_right[style_parameter['map_view_default_index']]
colorbar_data_one_slice_bokeh = ColumnDataSource(data=dict(colorbar_top=colorbar_top,colorbar_bottom=colorbar_bottom,\
colorbar_left=colorbar_data_one_slice['colorbar_left'],\
colorbar_right=colorbar_data_one_slice['colorbar_right'],\
palette_r=palette_r))
colorbar_data_all_slices_bokeh = ColumnDataSource(data=dict(colorbar_data_all_left=colorbar_data_all_left,\
colorbar_data_all_right=colorbar_data_all_right))
#
map_view_label_lon = style_parameter['map_view_depth_label_lon']
map_view_label_lat = style_parameter['map_view_depth_label_lat']
map_data_one_slice_depth = map_data_all_slices_depth[style_parameter['map_view_default_index']]
map_data_one_slice_depth_bokeh = ColumnDataSource(data=dict(lat=[map_view_label_lat], lon=[map_view_label_lon],
map_depth=[map_data_one_slice_depth]))
#
map_view_default_index = style_parameter['map_view_default_index']
map_data_one_slice = map_data_all_slices[map_view_default_index]
map_data_one_slice_bokeh = ColumnDataSource(data=dict(x=[style_parameter['map_view_image_lon_min']],\
y=[style_parameter['map_view_image_lat_min']],dw=[style_parameter['nlon']*style_parameter['dlon']],\
dh=[style_parameter['nlat']*style_parameter['dlat']],map_data_one_slice=[map_data_one_slice]))
map_data_all_slices_bokeh = ColumnDataSource(data=dict(map_data_all_slices=map_data_all_slices,\
map_data_all_slices_depth=map_data_all_slices_depth))
# ------------------------------
nprofile = len(profile_data_all)
grid_lat_list = []
grid_lon_list = []
width_list = []
height_list = []
for iprofile in range(nprofile):
aprofile = profile_data_all[iprofile]
grid_lat_list.append(aprofile['lat'])
grid_lon_list.append(aprofile['lon'])
width_list.append(style_parameter['map_view_grid_width'])
height_list.append(style_parameter['map_view_grid_height'])
grid_data_bokeh = ColumnDataSource(data=dict(lon=grid_lon_list,lat=grid_lat_list,\
width=width_list, height=height_list))
profile_default_index = style_parameter['profile_default_index']
selected_dot_on_map_bokeh = ColumnDataSource(data=dict(lat=[grid_lat_list[profile_default_index]], \
lon=[grid_lon_list[profile_default_index]], \
width=[style_parameter['map_view_grid_width']],\
height=[style_parameter['map_view_grid_height']],\
index=[profile_default_index]))
# ------------------------------
profile_vs_all = []
profile_depth_all = []
profile_ndepth = style_parameter['profile_ndepth']
profile_lat_label_list = []
profile_lon_label_list = []
for iprofile in range(nprofile):
aprofile = profile_data_all[iprofile]
vs_raw = aprofile['vs']
top_raw = aprofile['top']
profile_lat_label_list.append('Lat: {0:12.1f}'.format(aprofile['lat']))
profile_lon_label_list.append('Lon: {0:12.1f}'.format(aprofile['lon']))
vs_plot = []
depth_plot = []
for idepth in range(profile_ndepth):
vs_plot.append(vs_raw[idepth])
depth_plot.append(top_raw[idepth])
vs_plot.append(vs_raw[idepth])
depth_plot.append(top_raw[idepth+1])
profile_vs_all.append(vs_plot)
profile_depth_all.append(depth_plot)
profile_data_all_bokeh = ColumnDataSource(data=dict(profile_vs_all=profile_vs_all, \
profile_depth_all=profile_depth_all))
selected_profile_data_bokeh = ColumnDataSource(data=dict(vs=profile_vs_all[profile_default_index],\
depth=profile_depth_all[profile_default_index]))
selected_profile_lat_label_bokeh = ColumnDataSource(data=\
dict(x=[style_parameter['profile_lat_label_x']], y=[style_parameter['profile_lat_label_y']],\
lat_label=[profile_lat_label_list[profile_default_index]]))
selected_profile_lon_label_bokeh = ColumnDataSource(data=\
dict(x=[style_parameter['profile_lon_label_x']], y=[style_parameter['profile_lon_label_y']],\
lon_label=[profile_lon_label_list[profile_default_index]]))
all_profile_lat_label_bokeh = ColumnDataSource(data=dict(profile_lat_label_list=profile_lat_label_list))
all_profile_lon_label_bokeh = ColumnDataSource(data=dict(profile_lon_label_list=profile_lon_label_list))
#
button_ndepth = style_parameter['button_ndepth']
button_data_all_vs = []
button_data_all_vp = []
button_data_all_rho = []
button_data_all_top = []
for iprofile in range(nprofile):
aprofile = profile_data_all[iprofile]
button_data_all_vs.append(aprofile['vs'][:button_ndepth])
button_data_all_vp.append(aprofile['vp'][:button_ndepth])
button_data_all_rho.append(aprofile['rho'][:button_ndepth])
button_data_all_top.append(aprofile['top'][:button_ndepth])
button_data_all_bokeh = ColumnDataSource(data=dict(button_data_all_vs=button_data_all_vs,\
button_data_all_vp=button_data_all_vp,\
button_data_all_rho=button_data_all_rho,\
button_data_all_top=button_data_all_top))
# ==============================
map_view = Figure(plot_width=style_parameter['map_view_plot_width'], plot_height=style_parameter['map_view_plot_height'], \
tools=style_parameter['map_view_tools'], title=style_parameter['map_view_title'], \
y_range=[style_parameter['map_view_figure_lat_min'], style_parameter['map_view_figure_lat_max']],\
x_range=[style_parameter['map_view_figure_lon_min'], style_parameter['map_view_figure_lon_max']])
#
map_view.image('map_data_one_slice',x='x',\
y='y',dw='dw',\
dh='dh',palette=palette_r,\
source=map_data_one_slice_bokeh, level='image')
depth_slider_callback = CustomJS(args=dict(map_data_one_slice_bokeh=map_data_one_slice_bokeh,\
map_data_all_slices_bokeh=map_data_all_slices_bokeh,\
colorbar_data_all_slices_bokeh=colorbar_data_all_slices_bokeh,\
colorbar_data_one_slice_bokeh=colorbar_data_one_slice_bokeh,\
map_data_one_slice_depth_bokeh=map_data_one_slice_depth_bokeh), code="""
var d_index = Math.round(cb_obj.value)
var map_data_all_slices = map_data_all_slices_bokeh.data
map_data_one_slice_bokeh.data['map_data_one_slice'] = [map_data_all_slices['map_data_all_slices'][d_index]]
map_data_one_slice_bokeh.change.emit()
var color_data_all_slices = colorbar_data_all_slices_bokeh.data
colorbar_data_one_slice_bokeh.data['colorbar_left'] = color_data_all_slices['colorbar_data_all_left'][d_index]
colorbar_data_one_slice_bokeh.data['colorbar_right'] = color_data_all_slices['colorbar_data_all_right'][d_index]
colorbar_data_one_slice_bokeh.change.emit()
map_data_one_slice_depth_bokeh.data['map_depth'] = [map_data_all_slices['map_data_all_slices_depth'][d_index]]
map_data_one_slice_depth_bokeh.change.emit()
""")
depth_slider = Slider(start=0, end=style_parameter['map_view_ndepth']-1, \
value=map_view_default_index, step=1, \
width=style_parameter['map_view_plot_width'],\
title=style_parameter['depth_slider_title'], height=50, \
callback=depth_slider_callback)
# ------------------------------
# add boundaries to map view
# country boundaries
map_view.multi_line(boundary_data['country']['longitude'],\
boundary_data['country']['latitude'],color='black',\
line_width=2, level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# marine boundaries
map_view.multi_line(boundary_data['marine']['longitude'],\
boundary_data['marine']['latitude'],color='black',\
level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# shoreline boundaries
map_view.multi_line(boundary_data['shoreline']['longitude'],\
boundary_data['shoreline']['latitude'],color='black',\
line_width=2, level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# state boundaries
map_view.multi_line(boundary_data['state']['longitude'],\
boundary_data['state']['latitude'],color='black',\
level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# ------------------------------
# add depth label
map_view.rect(style_parameter['map_view_depth_box_lon'], style_parameter['map_view_depth_box_lat'], \
width=style_parameter['map_view_depth_box_width'], height=style_parameter['map_view_depth_box_height'], \
width_units='screen',height_units='screen', color='#FFFFFF', line_width=1., line_color='black', level='underlay')
map_view.text('lon', 'lat', 'map_depth', source=map_data_one_slice_depth_bokeh,\
text_font_size=style_parameter['annotating_text_font_size'],text_align='left',level='underlay')
# ------------------------------
map_view.rect('lon', 'lat', width='width', \
width_units='screen', height='height', \
height_units='screen', line_color='gray', line_alpha=0.5, \
selection_line_color='gray', selection_line_alpha=0.5, selection_fill_color=None,\
nonselection_line_color='gray',nonselection_line_alpha=0.5, nonselection_fill_color=None,\
source=grid_data_bokeh, color=None, line_width=1, level='glyph')
map_view.rect('lon', 'lat',width='width', \
width_units='screen', height='height', \
height_units='screen', line_color='#00ff00', line_alpha=1.0, \
source=selected_dot_on_map_bokeh, fill_color=None, line_width=3.,level='glyph')
# ------------------------------
grid_data_bokeh.callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh, \
grid_data_bokeh=grid_data_bokeh,\
profile_data_all_bokeh=profile_data_all_bokeh,\
selected_profile_data_bokeh=selected_profile_data_bokeh,\
selected_profile_lat_label_bokeh=selected_profile_lat_label_bokeh,\
selected_profile_lon_label_bokeh=selected_profile_lon_label_bokeh, \
all_profile_lat_label_bokeh=all_profile_lat_label_bokeh, \
all_profile_lon_label_bokeh=all_profile_lon_label_bokeh, \
), code="""
var inds = Math.round(cb_obj.selected['1d'].indices)
var grid_data = grid_data_bokeh.data
selected_dot_on_map_bokeh.data['lat'] = [grid_data['lat'][inds]]
selected_dot_on_map_bokeh.data['lon'] = [grid_data['lon'][inds]]
selected_dot_on_map_bokeh.data['index'] = [inds]
selected_dot_on_map_bokeh.change.emit()
var profile_data_all = profile_data_all_bokeh.data
selected_profile_data_bokeh.data['vs'] = profile_data_all['profile_vs_all'][inds]
selected_profile_data_bokeh.data['depth'] = profile_data_all['profile_depth_all'][inds]
selected_profile_data_bokeh.change.emit()
var all_profile_lat_label = all_profile_lat_label_bokeh.data['profile_lat_label_list']
var all_profile_lon_label = all_profile_lon_label_bokeh.data['profile_lon_label_list']
selected_profile_lat_label_bokeh.data['lat_label'] = [all_profile_lat_label[inds]]
selected_profile_lon_label_bokeh.data['lon_label'] = [all_profile_lon_label[inds]]
selected_profile_lat_label_bokeh.change.emit()
selected_profile_lon_label_bokeh.change.emit()
""")
# ------------------------------
# change style
map_view.title.text_font_size = style_parameter['title_font_size']
map_view.title.align = 'center'
map_view.title.text_font_style = 'normal'
map_view.xaxis.axis_label = style_parameter['map_view_xlabel']
map_view.xaxis.axis_label_text_font_style = 'normal'
map_view.xaxis.axis_label_text_font_size = xlabel_fontsize
map_view.xaxis.major_label_text_font_size = xlabel_fontsize
map_view.yaxis.axis_label = style_parameter['map_view_ylabel']
map_view.yaxis.axis_label_text_font_style = 'normal'
map_view.yaxis.axis_label_text_font_size = xlabel_fontsize
map_view.yaxis.major_label_text_font_size = xlabel_fontsize
map_view.xgrid.grid_line_color = None
map_view.ygrid.grid_line_color = None
map_view.toolbar.logo = None
map_view.toolbar_location = 'above'
map_view.toolbar_sticky = False
# ==============================
# plot colorbar
colorbar_fig = Figure(tools=[], y_range=(0,0.1),plot_width=style_parameter['map_view_plot_width'], \
plot_height=style_parameter['colorbar_plot_height'],title=style_parameter['colorbar_title'])
colorbar_fig.toolbar_location=None
colorbar_fig.quad(top='colorbar_top',bottom='colorbar_bottom',left='colorbar_left',right='colorbar_right',\
color='palette_r',source=colorbar_data_one_slice_bokeh)
colorbar_fig.yaxis[0].ticker=FixedTicker(ticks=[])
colorbar_fig.xgrid.grid_line_color = None
colorbar_fig.ygrid.grid_line_color = None
colorbar_fig.xaxis.axis_label_text_font_size = xlabel_fontsize
colorbar_fig.xaxis.major_label_text_font_size = xlabel_fontsize
colorbar_fig.xaxis[0].formatter = PrintfTickFormatter(format="%5.2f")
colorbar_fig.title.text_font_size = xlabel_fontsize
colorbar_fig.title.align = 'center'
colorbar_fig.title.text_font_style = 'normal'
# ==============================
profile_xrange = Range1d(start=style_parameter['profile_plot_xmin'], end=style_parameter['profile_plot_xmax'])
profile_yrange = Range1d(start=style_parameter['profile_plot_ymax'], end=style_parameter['profile_plot_ymin'])
profile_fig = Figure(plot_width=style_parameter['profile_plot_width'], plot_height=style_parameter['profile_plot_height'],\
x_range=profile_xrange, y_range=profile_yrange, tools=style_parameter['profile_tools'],\
title=style_parameter['profile_title'])
profile_fig.line('vs','depth',source=selected_profile_data_bokeh, line_width=2, line_color='black')
# ------------------------------
# add lat, lon
profile_fig.rect([style_parameter['profile_label_box_x']], [style_parameter['profile_label_box_y']],\
width=style_parameter['profile_label_box_width'], height=style_parameter['profile_label_box_height'],\
width_units='screen', height_units='screen', color='#FFFFFF', line_width=1., line_color='black',\
level='underlay')
profile_fig.text('x','y','lat_label', source=selected_profile_lat_label_bokeh)
profile_fig.text('x','y','lon_label', source=selected_profile_lon_label_bokeh)
# ------------------------------
# change style
profile_fig.xaxis.axis_label = style_parameter['profile_xlabel']
profile_fig.xaxis.axis_label_text_font_style = 'normal'
profile_fig.xaxis.axis_label_text_font_size = xlabel_fontsize
profile_fig.xaxis.major_label_text_font_size = xlabel_fontsize
profile_fig.yaxis.axis_label = style_parameter['profile_ylabel']
profile_fig.yaxis.axis_label_text_font_style = 'normal'
profile_fig.yaxis.axis_label_text_font_size = xlabel_fontsize
profile_fig.yaxis.major_label_text_font_size = xlabel_fontsize
profile_fig.xgrid.grid_line_dash = [4, 2]
profile_fig.ygrid.grid_line_dash = [4, 2]
profile_fig.title.text_font_size = style_parameter['title_font_size']
profile_fig.title.align = 'center'
profile_fig.title.text_font_style = 'normal'
profile_fig.toolbar_location = 'above'
profile_fig.toolbar_sticky = False
profile_fig.toolbar.logo = None
# ==============================
profile_slider_callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
grid_data_bokeh=grid_data_bokeh, \
profile_data_all_bokeh=profile_data_all_bokeh, \
selected_profile_data_bokeh=selected_profile_data_bokeh,\
selected_profile_lat_label_bokeh=selected_profile_lat_label_bokeh,\
selected_profile_lon_label_bokeh=selected_profile_lon_label_bokeh, \
all_profile_lat_label_bokeh=all_profile_lat_label_bokeh, \
all_profile_lon_label_bokeh=all_profile_lon_label_bokeh), code="""
var p_index = Math.round(cb_obj.value)
var grid_data = grid_data_bokeh.data
selected_dot_on_map_bokeh.data['lat'] = [grid_data['lat'][p_index]]
selected_dot_on_map_bokeh.data['lon'] = [grid_data['lon'][p_index]]
selected_dot_on_map_bokeh.data['index'] = [p_index]
selected_dot_on_map_bokeh.change.emit()
var profile_data_all = profile_data_all_bokeh.data
selected_profile_data_bokeh.data['vs'] = profile_data_all['profile_vs_all'][p_index]
selected_profile_data_bokeh.data['depth'] = profile_data_all['profile_depth_all'][p_index]
selected_profile_data_bokeh.change.emit()
var all_profile_lat_label = all_profile_lat_label_bokeh.data['profile_lat_label_list']
var all_profile_lon_label = all_profile_lon_label_bokeh.data['profile_lon_label_list']
selected_profile_lat_label_bokeh.data['lat_label'] = [all_profile_lat_label[p_index]]
selected_profile_lon_label_bokeh.data['lon_label'] = [all_profile_lon_label[p_index]]
selected_profile_lat_label_bokeh.change.emit()
selected_profile_lon_label_bokeh.change.emit()
""")
profile_slider = Slider(start=0, end=nprofile-1, value=style_parameter['profile_default_index'], \
step=1, title=style_parameter['profile_slider_title'], \
width=style_parameter['profile_plot_width'], height=50,\
callback=profile_slider_callback)
profile_slider_callback.args['profile_index'] = profile_slider
# ==============================
simple_text_button_callback = CustomJS(args=dict(button_data_all_bokeh=button_data_all_bokeh,\
selected_dot_on_map_bokeh=selected_dot_on_map_bokeh), \
code="""
var index = selected_dot_on_map_bokeh.data['index']
var button_data_vs = button_data_all_bokeh.data['button_data_all_vs'][index]
var button_data_vp = button_data_all_bokeh.data['button_data_all_vp'][index]
var button_data_rho = button_data_all_bokeh.data['button_data_all_rho'][index]
var button_data_top = button_data_all_bokeh.data['button_data_all_top'][index]
var csvContent = "data:text;charset=utf-8,"
var i = 0
var temp = csvContent
temp += "# Layer Top (km) Vs(km/s) Vp(km/s) Rho(g/cm^3) \\n"
while(button_data_vp[i]) {
temp+=button_data_top[i].toPrecision(6) + " " + button_data_vs[i].toPrecision(4) + " " + \
button_data_vp[i].toPrecision(4) + " " + button_data_rho[i].toPrecision(4) + "\\n"
i = i + 1
}
var encodedUri = encodeURI(temp)
link = document.createElement('a');
link.setAttribute('href', encodedUri);
link.setAttribute('download', 'vel_model.txt');
link.click();
""")
simple_text_button = Button(label=style_parameter['simple_text_button_label'], button_type='default', width=style_parameter['button_width'],\
callback=simple_text_button_callback)
# ------------------------------
model96_button_callback = CustomJS(args=dict(button_data_all_bokeh=button_data_all_bokeh,\
selected_dot_on_map_bokeh=selected_dot_on_map_bokeh), \
code="""
var index = selected_dot_on_map_bokeh.data['index']
var lat = selected_dot_on_map_bokeh.data['lat']
var lon = selected_dot_on_map_bokeh.data['lon']
var button_data_vs = button_data_all_bokeh.data['button_data_all_vs'][index]
var button_data_vp = button_data_all_bokeh.data['button_data_all_vp'][index]
var button_data_rho = button_data_all_bokeh.data['button_data_all_rho'][index]
var button_data_top = button_data_all_bokeh.data['button_data_all_top'][index]
var csvContent = "data:text;charset=utf-8,"
var i = 0
var temp = csvContent
temp += "MODEL." + index + " \\n"
temp += "ShearVelocityModel Lat: "+ lat +" Lon: " + lon + "\\n"
temp += "ISOTROPIC \\n"
temp += "KGS \\n"
temp += "SPHERICAL EARTH \\n"
temp += "1-D \\n"
temp += "CONSTANT VELOCITY \\n"
temp += "LINE08 \\n"
temp += "LINE09 \\n"
temp += "LINE10 \\n"
temp += "LINE11 \\n"
temp += " H(KM) VP(KM/S) VS(KM/S) RHO(GM/CC) QP QS ETAP ETAS FREFP FREFS \\n"
while(button_data_vp[i+1]) {
var thickness = button_data_top[i+1] - button_data_top[i]
temp+=" " +thickness.toPrecision(6) + " " + button_data_vp[i].toPrecision(4) + " " + button_data_vs[i].toPrecision(4) \
+ " " + button_data_rho[i].toPrecision(4) + " 0.00 0.00 0.00 0.00 1.00 1.00" + "\\n"
i = i + 1
}
var encodedUri = encodeURI(temp)
link = document.createElement('a');
link.setAttribute('href', encodedUri);
link.setAttribute('download', 'vel_model96.txt');
link.click();
""")
model96_button = Button(label=style_parameter['model96_button_label'], button_type='default', width=style_parameter['button_width'],\
callback=model96_button_callback)
# ==============================
# annotating text
annotating_fig01 = Div(text=style_parameter['annotating_html01'], \
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
annotating_fig02 = Div(text=style_parameter['annotating_html02'],\
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
# ==============================
output_file(filename,title=style_parameter['html_title'], mode=style_parameter['library_source'])
left_column = Column(depth_slider, map_view, colorbar_fig, annotating_fig01, width=style_parameter['left_column_width'])
button_pannel = Row(simple_text_button, model96_button)
right_column = Column(profile_slider, profile_fig, button_pannel, annotating_fig02, width=style_parameter['right_column_width'])
layout = Row(left_column, right_column)
save(layout)
def similarityGraph(si, sy, ey, db, abstractsim):
#calculate similarity and return necessary info
#ids = pmids of search, dates = dates of pmids, authors = authors of pmids, pmccites = number of cites, Y is x,y coordinate
#from dimensionality reduction, topwords = labels of clusters from kmeans/tfidf, kcenters = x,y centers of clusters from kmeans
#d is dictionary for all data storage
d = SimilarityCalc(si, sy, ey, db, abstractsim)
# for testing purposes
# with open('data.pk','wb') as f:
# pickle.dump(d,f)
#convert authors list of lists to list of strings for display
authors_str = []
for auths in d['authors']:
authors_str.append(", ".join(auths))
#calcualte a scaled pt size based on citation quantity
minw = 8
maxw = 30
# with open('citespkl.p','wb') as f:
# pickle.dump(list(map(int,pmccites)),f)
ptsizes = getScaledSizes(list(map(int, d['pmccites'])), minw, maxw)
#create colors based on years published
colors = getScaledColors(d['dates'])
#colors = ['blue']*len(ids)
alphas = [1] * len(d['ids'])
source = ColumnDataSource(data=dict(x=d['Y'][:, 0],
y=d['Y'][:, 1],
PMID=d['ids'],
titles=d['titles'],
authors=authors_str,
journals=d['journals'],
dates=d['dates'],
alphas=alphas,
pmccites=d['pmccites'],
ptsizes=ptsizes,
colors=colors,
colorsperm=colors))
########publication view table from selected on tsne plot
pubview_data = dict(titles=["Title"],
dates=["Date"],
journals=["Journal"],
authors=["Author"],
pmccites=["PMC Citations"],
PMID=["pmids"])
pubview_source = ColumnDataSource(pubview_data)
pubview_columns = [
TableColumn(field="titles", title="Article Title", width=400),
TableColumn(field="authors", title="Authors", width=50),
TableColumn(field="journals", title="Journal", width=50),
TableColumn(field="dates", title="Date", width=80),
TableColumn(field="pmccites", title="PMC Citations", width=80),
TableColumn(field="PMID", title="PMIDS", width=0),
]
pubview_table = DataTable(source=pubview_source,
columns=pubview_columns,
width=930,
height=400)
source.callback = CustomJS(args=dict(pubview_table=pubview_table),
code="""
var selecteddata = cb_obj.selected["1d"].indices
var count = 0
var s1 = cb_obj.get('data');
var d2 = pubview_table.get('source').get('data');
d2.index = []
d2.authors = []
d2.titles = []
d2.journals = []
d2.dates = []
d2.pmccites = []
d2.PMID = []
for(k = 0; k < selecteddata.length; k++){
tind = selecteddata[k]
d2.index.push(count)
d2.authors.push(s1.authors[tind])
d2.titles.push(s1.titles[tind])
d2.journals.push(s1.journals[tind])
d2.dates.push(s1.dates[tind])
d2.pmccites.push(parseInt(s1.pmccites[tind]))
d2.PMID.push(s1.PMID[tind])
count += 1
}
console.log(d2)
pubview_table.trigger('change');
""")
pubview_source.callback = CustomJS(code="""
var selecteddata = cb_obj.selected["1d"].indices
var s1 = cb_obj.get('data');
var url = "https://www.ncbi.nlm.nih.gov/pubmed/"+s1.PMID[selecteddata[0]]
window.open(url,'_blank');
""")
#######END TABLE DISPLAY CODE#####
#####max-width IS IMPORTANT FOR PROPER WRAPPING OF TEXT
hover = HoverTool(tooltips="""
<div>
<div style="max-width: 400px;">
<span style="font-size: 12px; font-weight: bold;">@titles</span>
</div>
<div style="max-width: 400px;">
<span style="font-size: 12px; color: #966;">@authors</span>
<div>
<div style="max-width: 400px;">
<span style="font-size: 12px; font-style: italic;">@journals, @dates</span>
<div style="max-width: 400px;">
<span style="font-size: 10px;">PMID</span>
<span style="font-size: 10px; color: #696;">@PMID</span>
</div>
<div style="max-width: 400px;">
<span style="font-size: 10px;">PMC Citations</span>
<span style="font-size: 10px; color: #696;">@pmccites</span>
</div>
</div>
""")
resetCallback = CustomJS(args=dict(source=source),
code="""
var data = source.get('data')
var titles = data['titles']
for (i=0; i < titles.length; i++) {
data.colors[i]=data.colorsperm[i]
data.alphas[i]= 1
}
source.trigger('change')
""")
#move function from reset callback to below so stuff updates automatically on textbox change
textCallback = CustomJS(args=dict(source=source),
code="""
var data = source.get('data')
var value = cb_obj.get('value')
var words = value.split(" ")
for (i=0; i < data.titles.length; i++) {
data.alphas[i]= 0.3
data.colors[i]=data.colorsperm[i]
}
for (i=0; i < data.titles.length; i++) {
for(j=0; j < words.length; j++){
if (data.titles[i].toLowerCase().indexOf(words[j].toLowerCase()) !== -1) {
if(j == words.length-1){
data.colors[i]='orange'
data.alphas[i]= 1
}
}else if(data.authors[i].toLowerCase().indexOf(words[j].toLowerCase()) !== -1){
if(j == words.length-1){
data.colors[i]='orange'
data.alphas[i]= 1
}
}else if(data.journals[i].toLowerCase().indexOf(words[j].toLowerCase()) !== -1){
if(j == words.length-1){
data.colors[i]='orange'
data.alphas[i]= 1
}
}else{
break
}
}
}
source.trigger('change')
""")
publistcallback = CustomJS(args=dict(pubview_table=pubview_table),
code="""
var pmids = pubview_table.get('source').get('data').PMID;
var pmidlist = pmids.join()
var url = "https://www.ncbi.nlm.nih.gov/pubmed/"+pmidlist
window.open(url,'_blank');
""")
TOOLS = 'pan,lasso_select,wheel_zoom,tap,reset'
p = figure(plot_width=900,
plot_height=600,
title="'" + si + "' tSNE similarity",
tools=[TOOLS, hover],
active_scroll='wheel_zoom',
active_drag="lasso_select")
p.circle('x',
'y',
fill_color='colors',
fill_alpha='alphas',
size='ptsizes',
line_color="#000000",
line_alpha=0.2,
source=source)
#word labeles for plots
wordsources = []
for idx in list(range(len(d['topwords']))):
wordsources.append(
ColumnDataSource(
dict(x=d['kcenters'][idx][:, 0],
y=d['kcenters'][idx][:, 1],
text=d['topwords'][idx])))
wordglyph = Text(x="x",
y="y",
text="text",
text_color="#000000",
text_font_style="bold",
text_font_size="14pt")
#5 is used for initial slider set below
initialclust = 5
wordholdsource = ColumnDataSource(
dict(x=d['kcenters'][initialclust][:, 0],
y=d['kcenters'][initialclust][:, 1],
text=d['topwords'][initialclust]))
p.add_glyph(wordholdsource, wordglyph)
# source = ColumnDataSource(data=dict(x=x, y=y))
#
# plot = Figure(plot_width=400, plot_height=400)
# plot.line('x', 'y', source=source, line_width=3, line_alpha=0.6)
args = {}
args["wordholdsource"] = wordholdsource
for idx in list(range(len(d['topwords']))):
args["wordsource" + str(idx + d['minc'])] = wordsources[idx]
#had to use eval hack because of limitations on the type of objects that can be passed into the callback, limited by bokeh backend
slidercallback = CustomJS(args=args,
code="""
var f = cb_obj.value
var ndata = eval('wordsource' + f.toString()).data;
wordholdsource.data.x = ndata.x
wordholdsource.data.y = ndata.y
wordholdsource.data.text = ndata.text
wordholdsource.trigger('change');
""")
wslider = Slider(start=d['minc'],
end=d['maxc'],
value=d['minc'] + initialclust,
step=1,
title="# of labels")
# slider = Slider(start=0.1, end=4, value=1, step=.1, title="power", callback=callback)
wslider.js_on_change('value', slidercallback)
#formatting plot
p.xaxis.axis_label = "Hover to view publication info, Click to open Pubmed link"
p.xaxis.major_tick_line_color = None # turn off x-axis major ticks
p.xaxis.minor_tick_line_color = None # turn off x-axis minor ticks
p.yaxis.major_tick_line_color = None # turn off y-axis major ticks
p.yaxis.minor_tick_line_color = None
p.xaxis.major_label_text_font_size = '0pt' # turn off x-axis tick labels
p.yaxis.major_label_text_font_size = '0pt'
left, right, bottom, top = np.amin(d['Y'][:, 0]) * 1.1, np.amax(
d['Y'][:, 0]) * 1.1, np.amin(d['Y'][:, 1]) * 1.1, np.amax(
d['Y'][:, 1]) * 1.1
p.x_range = Range1d(left, right)
p.y_range = Range1d(bottom, top)
#tap tool callback
url = "https://www.ncbi.nlm.nih.gov/pubmed/@PMID"
taptool = p.select(type=TapTool)
taptool.callback = OpenURL(url=url)
#work input callback
word_input = TextInput(title="Search for term(s) in graph",
placeholder="Enter term to highlight",
callback=textCallback)
reset = Button(label="Clear Highlighting",
callback=resetCallback,
width=150)
spdiv = Div(text=" ", width=100, height=20)
#add addition message saying data was cutoff in case of vague search terms
if (len(d['rids']) == d['maxids']):
cutoff_message = "<br>(Search was truncated to " + str(
d['maxids']) + " newest articles due to memory constraints)"
else:
cutoff_message = ""
tit1 = Div(text="<h1>" + si +
" similarity plot</h1><br><h5>Displaying the " +
str(d['sdfc_len']) + "/" + str(len(d['rids'])) +
" articles that have data on pubmed" + cutoff_message + "</h5>",
width=930)
pubmed_list_button = Button(label="Export selected publications to pubmed",
callback=publistcallback)
lt = layout([[tit1], [word_input], [reset, spdiv, wslider], [p],
[pubmed_list_button], [pubview_table]])
return lt
title="Testing",webgl=True)
p_voc = figure(tools=Tools_voc,
title="Testing",webgl=True,x_range = p_song.x_range,y_range = p_song.y_range)
p_song.circle('x', 'y', size='size',fill_alpha = 'alpha',fill_color='color',line_color=None, source=source_song)
p_voc.circle('x', 'y', size='size',fill_alpha = 'alpha',fill_color='color',line_color=None, source=source_voc)
source_song.callback = CustomJS(args=dict(source=source_voc), code="""
var inds = cb_obj.get('selected')['1d'].indices;
var data = source.get('data');
size = data['size'];
sorv = data['SorV'];
var selection = []
for (i=0;i<size.length;i++){
selection.push(0)
}
for (i=0;i<inds.length;i++){
selection[inds[i]]=1;
}
for (i=0;i<size.length;i++){
if(selection[i]==0&sorv[i]==1){
size[i]=0;
}
}
source.trigger('change');
""")
callback_alpha = CustomJS(args=dict(source_song=source_song,source_voc=source_voc), code="""
var song_data = source_song.get('data');
var voc_data = source_voc.get('data');
var f = cb_obj.get('value');
song_alpha = song_data['alpha'];
def selection_plot(response):
# Let's move these into a settings file somewhere?
# height/width could potentially be driven by the request?
# Include color data in the ColumnDataSource to allow for changing the color on
# the client side.
urls = [x[0] for x in response["pages"]]
xdata = [x[1] for x in response["pages"]]
ydata = [x[2] for x in response["pages"]]
tags = [x[3] for x in response["pages"]]
color = []
custom_tags = ["Custom tags"]
for tag in tags:
custom = False
if tag:
for t in tag:
if t not in ["Relevant", "Irrelevant", ""]:
custom = True
if t not in custom_tags:
custom_tags.append(t)
if not custom:
color.append(colormap(tag[0]))
else:
color.append(colormap("Custom"))
else:
color.append(colormap(None))
source = ColumnDataSource(
data=dict(
x=xdata,
y=ydata,
urls=urls,
tags=tags,
color=color,
)
)
# Callback code for CDS.
source.callback = CustomJS(code="""
var inds = cb_obj.get('selected')["1d"].indices;
var data = cb_obj.get('data');
BokehPlots.showPages(inds);
""")
# Create the figure with FIGURE_WIDTH and FIGURE_HEIGHT
p = figure(
tools="hover,wheel_zoom,reset",
width=FIGURE_WIDTH,
responsive=True,
tags=["clusterPlot"],
)
# Ensure that the lasso only selects with mouseup, not mousemove.
p.add_tools(
LassoSelectTool(select_every_mousemove=False),
)
# These turn off the x/y axis ticks
p.axis.visible = None
# These turn the major grid off
p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None
# Plot non-selected circles with a particular style using CIRCLE_SIZE and
# 'color' list
p.circle("x", "y", size=13, line_width=2, line_alpha=1,
line_color=None, fill_alpha=1, color='color', source=source,
name="urls")
nonselected_circle = Circle(fill_alpha=0.1, line_alpha=0.1, fill_color='color',
line_color='color')
renderer = p.select(name="urls")
renderer.nonselection_glyph = nonselected_circle
# Create buttons and their callbacks, use button_code string for callbacks.
button_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
tag = "%s";
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
selected: true,
possible: false,
});
data["color"][inds[i]] = "%s";
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
"""
textinput_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var tag = cb_obj.get("value");
// Reinitialise to the default value
cb_obj.set("value", "Add custom tag...")
if(tag.indexOf("Add custom tag...") < 0) {
//Update the custom tags selection list
var options = custom_tags_select.get("options");
if(options.indexOf(tag) < 0){
options.push(tag);
custom_tags_select.set("options", options);
}
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
selected: true,
possible: false,
});
data["color"][inds[i]] = "%s";
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
}
"""
selectinput_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var tag = cb_obj.get("value");
cb_obj.set("value", "Enter tags...")
if(tag.indexOf("Add custom tag...") < 0) {
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
selected: true,
possible: false,
});
data["color"][inds[i]] = "%s";
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
}
"""
# Create buttons and their callbacks, use button_code string for callbacks.
crawl_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var crawl = '%s';
for(var i = 0; i < inds.length; i++){
selected.push(data.urls[inds[i]]);
}
BokehPlots.crawlPages(selected, crawl);
source.trigger("change");
"""
# Supply color with print formatting.
but_relevant = Button(label="Relevant", type="success")
but_relevant.callback = CustomJS(args=dict(source=source),
code=button_code % ("Relevant", POSITIVE_COLOR))
but_irrelevant = Button(label="Irrelevant", type="success")
but_irrelevant.callback = CustomJS(args=dict(source=source),
code=button_code % ("Irrelevant", NEGATIVE_COLOR))
but_neutral = Button(label="Neutral", type="success")
but_neutral.callback = CustomJS(args=dict(source=source),
code=button_code % ("Neutral", NEUTRAL_COLOR))
custom_tag_input = TextInput(value="Add custom tag...")
custom_tag_input.callback = CustomJS(args=dict(source=source),
code=textinput_code % (CUSTOM_COLOR))
custom_tag_select = Select(value="Custom tags", options=custom_tags)
custom_tag_select.callback = CustomJS(args=dict(source=source),
code=selectinput_code % (CUSTOM_COLOR))
custom_tag_input.callback.args["custom_tags_select"] = custom_tag_select
but_backward_crawl = Button(label="Backlinks", type="success")
but_backward_crawl.callback = CustomJS(args=dict(source=source),
code=crawl_code % ("backward"))
but_forward_crawl = Button(label="Forwardlinks", type="success")
but_forward_crawl.callback = CustomJS(args=dict(source=source),
code=crawl_code % ("forward"))
# Adjust what attributes are displayed by the HoverTool
hover = p.select(dict(type=HoverTool))
hover.tooltips = [
("urls", "@urls"),
]
tags = hplot(custom_tag_input, custom_tag_select, but_neutral, but_relevant, but_irrelevant)
tags_crawl = hplot(but_backward_crawl, but_forward_crawl)
layout = vplot(p, tags, tags_crawl)
# Combine script and div into a single string.
plot_code = components(layout)
return plot_code[0] + plot_code[1]
def make_dashboard(motors, xy_train, operating_earnings, maintenance_cost, repair_cost, run_interval):
#calculate revenue vs time dataframe
print '...generating dashboard...'
events = get_events(motors)
events['earnings'] = 0.0
events.loc[events.state == 'operating', 'earnings'] = operating_earnings
events['expenses'] = 0.0
events.loc[events.state == 'maintenance', 'expenses'] = maintenance_cost
events.loc[events.state == 'repair', 'expenses'] = repair_cost
money = events.groupby('Time').sum()[['earnings', 'expenses']]
money['revenue'] = money.earnings - money.expenses
money['cumulative_earnings'] = money.earnings.cumsum()
money['cumulative_expenses'] = money.expenses.cumsum()
money['cumulative_revenue'] = money.revenue.cumsum()
#map the (P,T) decision surface
T_min = 50
T_max = 150
P_min = 0
P_max = 100
T_axis = np.arange(T_min, T_max, 0.5)
P_axis = np.arange(P_min, P_max, 0.5)
x, y = np.meshgrid(T_axis, P_axis)
ttf = np.zeros((len(P_axis), len(T_axis)))
import copy
m = copy.deepcopy(motors[0])
for p_idx in np.arange(len(P_axis)):
for t_idx in np.arange(len(T_axis)):
m.Temp = T_axis[t_idx]
m.Pressure = P_axis[p_idx]
ttf[p_idx, t_idx] = m.predicted_time_to_fail()
#plot decision surface
from bokeh.plotting import figure, show, output_file, ColumnDataSource, vplot
from bokeh.models import HoverTool, Callback
from bokeh.io import vform
output_file('dashboard.html', title='Smart Maintenance Dashboard')
source = ColumnDataSource(
data=dict(
x = xy_train.Temp,
y = xy_train.Pressure,
ttf = xy_train.time_to_fail,
size = 0.6*xy_train.time_to_fail,
)
)
dec_fig = figure(x_range=[T_min, T_max], y_range=[P_min, P_max], title='SVM Decision Surface',
x_axis_label='Temperature', y_axis_label='Pressure', tools='box_zoom,reset,hover,crosshair',
width=600, plot_height=600)
dec_fig.title_text_font_size = '18pt'
dec_fig.xaxis.axis_label_text_font_size = '14pt'
dec_fig.yaxis.axis_label_text_font_size = '14pt'
dec_fig.image(image=[-ttf], x=[T_min], y=[P_min], dw=[T_max - T_min], dh=[P_max - P_min],
palette='RdYlGn8')
dec_fig.x('x', 'y', size='size', source=source, fill_alpha=0.5, fill_color='navy',
line_color='navy', line_width=1, line_alpha=0.5)
hover = dec_fig.select(dict(type=HoverTool))
hover.tooltips = [
("Temperature", "@x"),
("Pressure", "@y"),
("measured lifetime", "@ttf"),
]
#plot earnings vs time
source = ColumnDataSource(
data=dict(
t = money.index,
earnings = money.cumulative_earnings/1.e6,
expenses = money.cumulative_expenses/1.e6,
revenue = money.cumulative_revenue/1.e6,
zero = money.cumulative_revenue*0,
)
)
earn_fig = figure(title='Cumulative Earnings & Expenses', x_axis_label='Time',
y_axis_label='Earnings & Expenses (M$)', tools='box_zoom,reset,hover,crosshair',
width=1000, plot_height=300, x_range=[0, 1200], y_range=[0, 120])
earn_fig.title_text_font_size = '15pt'
earn_fig.xaxis.axis_label_text_font_size = '11pt'
earn_fig.yaxis.axis_label_text_font_size = '11pt'
earn_fig.line('t', 'earnings', color='blue', source=source, line_width=5, legend='earnings')
earn_fig.line('t', 'expenses', color='red', source=source, line_width=5, legend='expenses')
earn_fig.legend.orientation = "bottom_right"
earn_fig.patch([0, 200, 200, 0], [0, 0, 120, 120], color='lightsalmon', alpha=0.35,
line_width=0)
earn_fig.patch([200, 400, 400, 200], [0, 0, 120, 120], color='gold', alpha=0.35,
line_width=0)
earn_fig.patch([400, 1200, 1200, 400], [0, 0, 120, 120], color='darkseagreen',
alpha=0.35, line_width=0)
earn_fig.text([45], [101], ['run-to-fail'])
earn_fig.text([245], [101], ['scheduled'])
earn_fig.text([245], [90], ['maintenance'])
earn_fig.text([445], [101], ['predictive'])
earn_fig.text([445], [90], ['maintenance'])
hover = earn_fig.select(dict(type=HoverTool))
hover.tooltips = [
(" Time", "@t"),
(" earning (M$)", "@earnings"),
("expenses (M$)", "@expenses"),
]
#plot revenue vs time
rev_fig = figure(title='Cumulative Revenue', x_axis_label='Time',
y_axis_label='Revenue (M$)', tools='box_zoom,reset,hover,crosshair',
width=1000, plot_height=300, x_range=[0, 1200], y_range=[-15, 10])
rev_fig.title_text_font_size = '15pt'
rev_fig.xaxis.axis_label_text_font_size = '11pt'
rev_fig.yaxis.axis_label_text_font_size = '11pt'
rev_fig.line('t', 'revenue', color='green', source=source, line_width=5, legend='revenue')
rev_fig.line('t', 'zero', color='purple', source=source, line_width=3, alpha=0.5,
line_dash=[10, 5])
rev_fig.legend.orientation = "bottom_right"
rev_fig.patch([0, 200, 200, 0], [-15, -15, 10, 10], color='lightsalmon', alpha=0.35,
line_width=0)
rev_fig.patch([200, 400, 400, 200], [-15, -15, 10, 10], color='gold', alpha=0.35,
line_width=0)
rev_fig.patch([400, 1200, 1200, 400], [-15, -15, 10, 10], color='darkseagreen',
alpha=0.35, line_width=0)
hover = rev_fig.select(dict(type=HoverTool))
hover.tooltips = [
(" Time", "@t"),
(" revenue (M$)", "@revenue"),
]
#plot number of motors vs time
N = events.groupby(['Time', 'state']).count().unstack()['id'].reset_index()
N.fillna(value=0, inplace=True)
N['total'] = N.maintenance + N.operating + N.repair
s1 = ColumnDataSource(
data=dict(
Time = N.Time,
operating = N.operating,
maintenance = N.maintenance,
repair = N.repair,
total = N.total,
)
)
motor_fig = figure(title='Number of Motors', x_axis_label='Time',
y_axis_label='Number of motors', tools='box_zoom,reset,hover,crosshair',
width=1000, plot_height=300, x_range=[0, 1200], y_range=[-10, 210])
motor_fig.title_text_font_size = '15pt'
motor_fig.xaxis.axis_label_text_font_size = '11pt'
motor_fig.yaxis.axis_label_text_font_size = '11pt'
motor_fig.line('Time', 'total', color='blue', source=s1, line_width=3, legend='total')
motor_fig.line('Time', 'operating', color='green', source=s1, line_width=3, legend='operating')
motor_fig.line('Time', 'maintenance', color='orange', source=s1, line_width=3, legend='maintenance')
motor_fig.line('Time', 'repair', color='red', source=s1, line_width=3, legend='repair')
motor_fig.legend.orientation = "top_right"
motor_fig.patch([0, 200, 200, 0], [-10, -10, 210, 210], color='lightsalmon', alpha=0.35,
line_width=0)
motor_fig.patch([200, 400, 400, 200], [-10, -10, 210, 210], color='gold', alpha=0.35,
line_width=0)
motor_fig.patch([400, 1200, 1200, 400], [-10, -10, 210, 210], color='darkseagreen',
alpha=0.35, line_width=0)
#display N table
from bokeh.models.widgets import DataTable, TableColumn
from bokeh.io import vform
columns = [
TableColumn(field='Time', title='Time'),
TableColumn(field='operating', title='operating'),
TableColumn(field='maintenance', title='maintenance'),
TableColumn(field='repair', title='repair'),
TableColumn(field='total', title='total'),
]
s2 = s1.clone()
N_table = DataTable(source=s2, columns=columns, width=600, height=300)
s1.callback = Callback(args=dict(s2=s2), code="""
var inds = cb_obj.get('selected')['1d'].indices;
var d1 = cb_obj.get('data');
var d2 = s2.get('data');
d2['Time'] = []
d2['operating'] = []
d2['maintenance'] = []
d2['repair'] = []
d2['total'] = []
for (i = 0; i < inds.length; i++) {
d2['Time'].push(d1['Time'][inds[i]])
d2['operating'].push(d1['operating'][inds[i]])
d2['maintenance'].push(d1['maintenance'][inds[i]])
d2['repair'].push(d1['repair'][inds[i]])
d2['total'].push(d1['total'][inds[i]])
}
s2.trigger('change');
""")
#export plot to html and return
plot_grid = vplot(dec_fig, earn_fig, rev_fig, motor_fig, vform(N_table))
show(plot_grid, new='tab')
return money, events,N
#First plot
s1 = ColumnDataSource(data=dict(x=x, y=y))
p1 = figure(tools=["lasso_select"], plot_width=600, plot_height=400)
p1.scatter('x', 'y', fill_color='black', line_color=None, size=10, source=s1)
#Second plot
s2 = ColumnDataSource(data=dict(x=[], y=[], y2=[]))
p2 = figure(plot_width=400, plot_height=400, tools=[])
m1 = ColumnDataSource(m1) #Actual Datasource for the second plot
p2.line(np.arange(0, 100, 1), 'y', source=s2) # From original data - series 1
p2.line(np.arange(0, 100, 1), 'y2', source=s2) # From original data - series 2
s1.callback = Callback(args=dict(s2=s2, m1=m1),
code="""
var inds = cb_obj.get('selected')['1d'].indices;
var d1 = m1.get('data');
var d2 = s2.get('data');
d2['y'] = []
d2['y2'] = []
for (i = 0; i < 11; i++) {o
d2['y'].push(d1[inds['0']][i]),
d2['y2'].push(d1[inds['1']][i])
}
s2.trigger('change');
""")
layout = hplot(p1, p2)
show(layout)
plot_source2 = ColumnDataSource(data=dict(x=[], y=[]))
p2 = figure(plot_width=500, x_range=(0, 1), y_range=(0, 1))
p2.circle('x', 'y', source=plot_source2)
# scr2, di2 = components(p)
p3 = layouts.column(p, p2)
scr, di = components(p3)
plot_source.callback = CustomJS(args=dict(s2=plot_source2),
code="""
var inds = cb_obj.selected['1d'].indices;
var d1 = cb_obj.data;
var d2 = s2.data;
d2['x'] = []
d2['y'] = []
for (i=0; i<inds.length; i++) {
d2['x'].push(d1['x'][inds[i]])
d2['y'].push(d1['y'][inds[j]])
}
s2.trigger('change');
""")
# plot_source.on_change('selected',flash_selected)
###############################################################
# given tweet index in tweet_list, return the tweet blockquote embed code
def get_tweet_blockquote(tweet_index):
tweet_id = tweet_list[tweet_index]['id_str']
def selection_plot(response, tag_colors):
# Let's move these into a settings file somewhere?
# height/width could potentially be driven by the request?
# Include color data in the ColumnDataSource to allow for changing the color on
# the client side.
urls = [x[0] for x in response["pages"]]
xdata = [x[1] for x in response["pages"]]
ydata = [x[2] for x in response["pages"]]
tags = [x[3] for x in response["pages"]]
color = []
custom_tags = ["Custom tags"]
for tag in tags:
if tag:
t = tag[len(tag) - 1]
if t not in ["Relevant", "Irrelevant", ""]:
if t not in custom_tags:
custom_tags.append(t)
if ((tag_colors != None) and (t in tag_colors["colors"])):
color.append(tag_colors["colors"][t])
else:
color.append(colormap("Custom"))
else:
color.append(colormap(t))
else:
color.append(colormap(None))
if tag_colors != None:
for tag in tag_colors["colors"].keys():
if tag not in custom_tags:
custom_tags.append(tag)
source = ColumnDataSource(data=dict(x=xdata,
y=ydata,
urls=urls,
tags=tags,
color=color,
custom_tags=custom_tags))
# Callback code for CDS.
source.callback = CustomJS(code="""
var inds = cb_obj.get('selected')["1d"].indices;
var data = cb_obj.get('data');
BokehPlots.addCustomTags(data['custom_tags']);
BokehPlots.showPages(inds);
""")
# Create the figure with FIGURE_WIDTH and FIGURE_HEIGHT
p = figure(
tools="hover,wheel_zoom,reset",
width=FIGURE_WIDTH,
height=FIGURE_HEIGHT,
responsive=True,
tags=["clusterPlot"],
min_border_bottom=MIN_BORDER_BOTTOM,
min_border_top=MIN_BORDER_TOP,
min_border_left=MIN_BORDER_LEFT,
min_border_right=MIN_BORDER_RIGHT,
)
# Ensure that the lasso only selects with mouseup, not mousemove.
p.add_tools(LassoSelectTool(select_every_mousemove=False), )
# These turn off the x/y axis ticks
p.axis.visible = None
# These turn the major grid off
p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None
# Plot non-selected circles with a particular style using CIRCLE_SIZE and
# 'color' list
p.circle("x",
"y",
size=13,
line_width=2,
line_alpha=1,
line_color=None,
fill_alpha=1,
color='color',
source=source,
name="urls")
nonselected_circle = Circle(fill_alpha=0.1,
line_alpha=0.1,
fill_color='color',
line_color='color')
renderer = p.select(name="urls")
renderer.nonselection_glyph = nonselected_circle
# Create buttons and their callbacks, use button_code string for callbacks.
button_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
tag = "%s";
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
color: data.color[inds[i]],
selected: true,
possible: false,
});
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
"""
textinput_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var tag = cb_obj.get("value");
// Reinitialise to the default value
cb_obj.set("value", "Add custom tag...")
if(tag.indexOf("Add custom tag...") < 0) {
//Update the custom tags selection list
var options = custom_tags_select.get("options");
if(options.indexOf(tag) < 0){
options.push(tag);
custom_tags_select.set("options", options);
}
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
color: data.color[inds[i]],
selected: true,
possible: false,
});
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
}
"""
selectinput_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var tag = cb_obj.get("value");
cb_obj.set("value", "Enter tags...")
if(tag.indexOf("Add custom tag...") < 0) {
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
color: data.color[inds[i]],
selected: true,
possible: false,
});
}
BokehPlots.updateTags(selected, tag, "Apply");
source.trigger("change");
}
"""
# Create buttons and their callbacks, use button_code string for callbacks.
crawl_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
var crawl = '%s';
for(var i = 0; i < inds.length; i++){
selected.push(data.urls[inds[i]]);
}
BokehPlots.crawlPages(selected, crawl);
source.trigger("change");
"""
# Supply color with print formatting.
but_relevant = Button(label="Relevant", type="success")
but_relevant.callback = CustomJS(args=dict(source=source),
code=button_code % ("Relevant"))
but_irrelevant = Button(label="Irrelevant", type="success")
but_irrelevant.callback = CustomJS(args=dict(source=source),
code=button_code % ("Irrelevant"))
but_neutral = Button(label="Neutral", type="success")
but_neutral.callback = CustomJS(args=dict(source=source),
code=button_code % ("Neutral"))
custom_tag_input = TextInput(value="Add custom tag...")
custom_tag_input.callback = CustomJS(args=dict(source=source),
code=textinput_code % ())
custom_tag_select = Select(value="Custom tags", options=custom_tags)
custom_tag_select.callback = CustomJS(args=dict(source=source),
code=selectinput_code % ())
custom_tag_input.callback.args["custom_tags_select"] = custom_tag_select
but_backward_crawl = Button(label="Backlinks", type="success")
but_backward_crawl.callback = CustomJS(args=dict(source=source),
code=crawl_code % ("backward"))
but_forward_crawl = Button(label="Forwardlinks", type="success")
but_forward_crawl.callback = CustomJS(args=dict(source=source),
code=crawl_code % ("forward"))
# Adjust what attributes are displayed by the HoverTool
hover = p.select(dict(type=HoverTool))
hover.tooltips = [
("urls", "@urls"),
]
tags = hplot(but_neutral,
but_relevant,
but_irrelevant,
custom_tag_input,
custom_tag_select,
height=40)
tags_crawl = hplot(but_backward_crawl, but_forward_crawl)
layout = vplot(p, tags, tags_crawl)
# Combine script and div into a single string.
plot_code = components(layout)
return plot_code[0] + plot_code[1]
def selection_plot(response):
# Let's move these into a settings file somewhere?
# height/width could potentially be driven by the request?
# Include color data in the ColumnDataSource to allow for changing the color on
# the client side.
urls = [x[0] for x in response["pages"]]
xdata = [x[1] for x in response["pages"]]
ydata = [x[2] for x in response["pages"]]
tags = [x[3] for x in response["pages"]]
color = [colormap(x[0]) if x else colormap(None) for x in tags]
source = ColumnDataSource(
data=dict(
x=xdata,
y=ydata,
urls=urls,
tags=tags,
color=color,
)
)
# Callback code for CDS.
source.callback = CustomJS(code="""
var inds = cb_obj.get('selected')["1d"].indices;
var data = cb_obj.get('data');
BokehPlots.showPages(inds);
""")
# Create the figure with FIGURE_WIDTH and FIGURE_HEIGHT
p = figure(tools="hover", width=FIGURE_WIDTH,
toolbar_location=None, responsive=True, tags=["clusterPlot"])
# Ensure that the lasso only selects with mouseup, not mousemove.
p.add_tools(LassoSelectTool(select_every_mousemove=False))
# These turn off the x/y axis ticks
p.axis.visible = None
# These turn the major grid off
p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None
# Plot non-selected circles with a particular style using CIRCLE_SIZE and
# 'color' list
p.circle("x", "y", size=13, line_width=2, line_alpha=1,
line_color=None, fill_alpha=1, color='color', source=source,
name="urls")
nonselected_circle = Circle(fill_alpha=0.1, line_alpha=0.1, fill_color='color',
line_color='color')
renderer = p.select(name="urls")
renderer.nonselection_glyph = nonselected_circle
# Create buttons and their callbacks, use button_code string for callbacks.
button_code = """
event.preventDefault();
var inds = source.get('selected')["1d"].indices;
var data = source.get('data');
var selected = [];
tag = "%s";
for(var i = 0; i < inds.length; i++){
selected.push({
x: data.x[inds[i]],
y: data.y[inds[i]],
url: data.urls[inds[i]],
tags: data.tags[inds[i]],
selected: true,
possible: false,
});
data["color"][inds[i]] = "%s";
}
BokehPlots.updateTags(selected, tag, inds);
source.trigger("change");
"""
# Supply color with print formatting.
button1 = Button(label="Relevant", type="success")
button1.callback = CustomJS(args=dict(source=source),
code=button_code % ("Relevant", POSITIVE_COLOR))
button2 = Button(label="Irrelevant", type="success")
button2.callback = CustomJS(args=dict(source=source),
code=button_code % ("Irrelevant", NEGATIVE_COLOR))
button3 = Button(label="Neutral", type="success")
button3.callback = CustomJS(args=dict(source=source),
code=button_code % ("Neutral", NEUTRAL_COLOR))
# Adjust what attributes are displayed by the HoverTool
hover = p.select(dict(type=HoverTool))
hover.tooltips = [
("urls", "@urls"),
]
layout = vform(p, button1, button2, button3)
# Combine script and div into a single string.
plot_code = components(layout)
return plot_code[0] + plot_code[1]
source_song.callback = CustomJS(
args=dict(source=source_voc, sf=source_voc_fix, svs=source_voc_size),
code="""
var inds = cb_obj.get('selected')['1d'].indices;
var song_data = cb_obj.get('data');
var voc_data = source.get('data');
var voc_size = sf.get('data')['size'][0];
var voc_alpha = sf.get('data')['alpha'][0];
var voc_size_tmp = svs.get('data')['size'];
function sum( obj ) {
var sum = 0;
for( var el in obj ) {
if( obj.hasOwnProperty( el ) ) {
sum += parseFloat( obj[el] );
}
}
return sum;
}
function normalize(voc){
var summed = sum(voc);
//console.log(Object.keys(voc));
for(var el in voc){
voc[el] = voc[el]/summed;
}
return voc;
}
Object.extend = function(destination, source) {
for (var property in source) {
if (destination.hasOwnProperty(property)) {
destination[property] += source[property];
}else{
destination[property]=source[property];
}
}
return destination;
};
var vocs = {
};
for (i = 0; i < inds.length; i++) {
vocs = Object.extend(vocs, song_data['lyrics'][inds[i]]);
//var key = Object.keys(song_data['lyrics'][inds[i]]);
//for (j=0;j<key.length;j++){
// vocs.add(key[j]);
//}
}
var voc_sum = sum(vocs);
vocs = normalize(vocs);//do the normalization to all voc
var key = Object.keys(vocs);
var select_array = new Array(voc_data['alpha'].length);
key.forEach(function(value) {
select_array[value]=1;
});
function componentToHex(c) {
var hex = c.toString(16);
return hex.length == 1 ? "0" + hex : hex;
}
function rgbToHex(r, g, b) {
return "#" + componentToHex(r) + componentToHex(g) + componentToHex(b);
}
for (i=0;i<select_array.length;i++){
if(inds.length==0){
voc_data['alpha'][i] = voc_alpha;
voc_data['size'][i] = voc_size_tmp[i];//TODO:Change to Document Freq.
voc_data['color'][i] = rgbToHex(255, 0, 0);
}else{
if(select_array[i]==1){
voc_data['alpha'][i] = voc_alpha;
voc_data['size'][i] = Math.pow(vocs[i]*voc_data['docfreq'][i]*select_array.length/inds.length,voc_size);//TODO:Change to tf-idf with normalized voc tf with normalized by song count
//voc_data['size'][i] = Math.pow(vocs[i]*select_array.length/inds.length,voc_size);//TODO:Change to tf
//voc_data['size'][i] = Math.pow(vocs[i]*voc_data['docfreq'][i]*voc_sum,voc_size);//TODO:Change to tf-idf with normalized voc tf with normalized by term count
//voc_data['size'][i] = Math.pow(vocs[i]*voc_sum,voc_size);//TODO:Change to tf
voc_data['color'][i] = rgbToHex(255,voc_data['size'][i],0);//map from size
voc_size_tmp[i]=voc_data['size'][i];
}else{
voc_data['alpha'][i] = 0;
voc_data['size'][i] = 0;
voc_data['color'][i] = rgbToHex(255, 0, 0);
}
}
}
source.trigger('change');
""",
)
def yearGraph(si, sy, ey):
years = list(range(sy, ey))
ss = quote(si)
p = figure(title="Articles containing: " + si,
plot_width=600,
plot_height=300,
x_axis_label="Year",
y_axis_label="Total Articles")
##FOR SEARCHING YOU WILL NEED TO ESCAPE SPACES AND SPECIAL CHARS
yearcounts, ids = getYears(ss, years)
pre = time.time()
titles, dates, authors, journals, pmccites = getPMIDInfo(ids)
print("Gathered PMID Summary for counts:" + str(time.time() - pre))
p.line(years, yearcounts, color='blue')
#Parse out statisitcs about returned data
authors_str = []
for auths in authors:
authors_str.append(", ".join(auths))
pmidPD = pd.DataFrame(
data={
'titles': titles,
'dates': dates,
'authors': authors,
'authors_str': authors_str,
'journals': journals,
'pmccites': list(map(int, pmccites))
})
#print(pmidPD[:5])
topPmids = pmidPD.nlargest(10, 'pmccites')
#####ADD CODE TO DISPLAY TABLE NEXT TO PLOT, determine how good a proxy pmccites is for real number of citation
#print(topPmids)
total_authors = Counter(auth for auths in pmidPD['authors']
for auth in auths)
total_authors_norm = Counter()
for idx, auths in enumerate(pmidPD['authors']):
for auth in auths:
total_authors_norm[auth] += pmidPD['pmccites'][idx]
total_journals = Counter(jour for jour in pmidPD['journals'])
#get data ready for display in bokeh
#display nent number of entries
nent = 100
topauths = total_authors.most_common(nent)
topauths_norm = total_authors_norm.most_common(nent)
topjournals = total_journals.most_common(nent)
#for displaying authors based on citations and papers
paper_data = dict(numpapers=[n[1] for n in topauths],
auths=[n[0] for n in topauths])
cite_data = dict(numcites=[n[1] for n in topauths_norm],
authscites=[n[0] for n in topauths_norm])
journal_data = dict(numpapers=[n[1] for n in topjournals],
journals=[n[0] for n in topjournals])
#print(len(titles),len(dates),len(journals),len(authors_str),len(pmccites))
#print(journals)
pub_data = dict(titles=titles,
dates=dates,
journals=journals,
authors=authors_str,
pmccites=pmccites,
PMID=ids)
pubview_data = dict(titles=["Title"],
dates=["Date"],
journals=["Journal"],
authors=["Author"],
pmccites=["PMC Citations"],
PMID=["PMID"])
#publication table that will be populated by
paper_source = ColumnDataSource(data=paper_data)
cite_source = ColumnDataSource(data=cite_data)
journal_source = ColumnDataSource(data=journal_data)
pub_source = ColumnDataSource(pub_data)
pubview_source = ColumnDataSource(pubview_data)
paper_columns = [
TableColumn(field="numpapers", title="# Papers", width=100),
TableColumn(field="auths", title="Author", width=150),
]
cite_columns = [
TableColumn(field="numcites", title="# PMC Citations", width=130),
TableColumn(field="authscites", title="Author", width=150),
]
journal_columns = [
TableColumn(field="numpapers", title="# Papers", width=80),
TableColumn(field="journals", title="Journal", width=220),
]
template = """
<div style="background:red
color: white">
<%= value %>\ninfo</div>
"""
formatter = HTMLTemplateFormatter(template=template)
pubview_columns = [
TableColumn(field="titles", title="Article Title",
formatter=formatter),
TableColumn(field="authors", title="Authors"),
TableColumn(field="journals", title="Journal"),
TableColumn(field="dates", title="Date", width=80),
TableColumn(field="pmccites", title="PMC Citations", width=80),
]
paper_table = DataTable(source=paper_source,
columns=paper_columns,
width=250,
height=210)
cite_table = DataTable(source=cite_source,
columns=cite_columns,
width=280,
height=210)
journal_table = DataTable(source=journal_source,
columns=journal_columns,
width=300,
height=210)
tit0 = Div(text="<h1>General statistics for " + si + "</h1>", width=930)
tit1 = Div(
text=
"<h1>Common Authors and Journals</h1>\n<h5>(Select author(s) or journal(s) to display related publications)</h5>",
width=930)
tit2 = Div(
text=
"<h1>Related Publications</h1>\n<h5>(click publication to visit pubmed entry)</h5>",
width=930)
spacer = figure(plot_width=50,
plot_height=210,
logo=None,
toolbar_location=None,
outline_line_color=None)
spacer2 = figure(plot_width=50,
plot_height=210,
logo=None,
toolbar_location=None,
outline_line_color=None)
#create table updated by clicks in above tables, for viewing related publications
pubview_table = DataTable(source=pubview_source,
columns=pubview_columns,
width=930,
height=400)
# callbacks for each of the table selections
#JScallback for setting up publications table
#curspot
paper_source.callback = CustomJS(args=dict(paper_source=paper_source,
pub_source=pub_source,
pubview_table=pubview_table),
code="""
var authordata = paper_source.selected["1d"].indices
var author = 'test'
var count = 0
var s1 = paper_source.get('data');
var d1 = pub_source.get('data');
var d2 = pubview_table.get('source').get('data');
d2.index = []
d2.authors = []
d2.titles = []
d2.journals = []
d2.dates = []
d2.pmccites = []
d2.PMID = []
for(j = 0; j < d1.authors.length; j++){
for(k = 0; k < authordata.length; k++){
if (d1.authors[j].toLowerCase().indexOf(s1.auths[authordata[k]].toLowerCase()) !== -1) {
d2.index.push(count)
d2.authors.push(d1.authors[j])
d2.titles.push(d1.titles[j])
d2.journals.push(d1.journals[j])
d2.dates.push(d1.dates[j])
d2.pmccites.push(parseInt(d1.pmccites[j]))
d2.PMID.push(d1.PMID[j])
count += 1
break;
}
}
}
console.log(d2)
pubview_table.trigger('change');
""")
cite_source.callback = CustomJS(args=dict(cite_source=cite_source,
pub_source=pub_source,
pubview_table=pubview_table),
code="""
var authordata = cite_source.selected["1d"].indices
var author = 'test'
var count = 0
var s1 = cite_source.get('data');
var d1 = pub_source.get('data');
var d2 = pubview_table.get('source').get('data');
d2.index = []
d2.authors = []
d2.titles = []
d2.journals = []
d2.dates = []
d2.pmccites = []
d2.PMID = []
for(j = 0; j < d1.authors.length; j++){
for(k = 0; k < authordata.length; k++){
if (d1.authors[j].toLowerCase().indexOf(s1.authscites[authordata[k]].toLowerCase()) !== -1) {
d2.index.push(count)
d2.authors.push(d1.authors[j])
d2.titles.push(d1.titles[j])
d2.journals.push(d1.journals[j])
d2.dates.push(d1.dates[j])
d2.pmccites.push(parseInt(d1.pmccites[j]))
d2.PMID.push(d1.PMID[j])
count += 1
break;
}
}
}
console.log(d2)
pubview_table.trigger('change');
""")
journal_source.callback = CustomJS(args=dict(journal_source=journal_source,
pub_source=pub_source,
pubview_table=pubview_table),
code="""
var journaldata = journal_source.selected["1d"].indices
var author = 'test'
var count = 0
var s1 = journal_source.get('data');
var d1 = pub_source.get('data');
var d2 = pubview_table.get('source').get('data');
d2.index = []
d2.authors = []
d2.titles = []
d2.journals = []
d2.dates = []
d2.pmccites = []
d2.PMID = []
for(j = 0; j < d1.journals.length; j++){
for(k = 0; k < journaldata.length; k++){
//checks to make sure journal name exists, seems like certain publications have no journal name in pubmed
if(d1.journals[j]){
if (d1.journals[j].toLowerCase() == s1.journals[journaldata[k]].toLowerCase()) {
d2.index.push(count)
d2.authors.push(d1.authors[j])
d2.titles.push(d1.titles[j])
d2.journals.push(d1.journals[j])
d2.dates.push(d1.dates[j])
d2.pmccites.push(parseInt(d1.pmccites[j]))
d2.PMID.push(d1.PMID[j])
count += 1
break;
}
}
}
}
console.log(d2)
pubview_table.trigger('change');
""")
pubview_source.callback = CustomJS(code="""
var selecteddata = cb_obj.selected["1d"].indices
var s1 = cb_obj.get('data');
var url = "https://www.ncbi.nlm.nih.gov/pubmed/"+s1.PMID[selecteddata[0]]
window.open(url,'_blank');
""")
###### write callbacks for other two tables as well
yearslayout = layout(
[[tit0], [p], [tit1],
[paper_table, spacer, cite_table, spacer2, journal_table], [tit2],
[pubview_table]])
#show(yearslayout)
# print(total_authors.most_common(10))
# print(total_authors_norm.most_common(10))
# print(total_authors['Harley CB'], total_authors['Thomson JA'])
#show(p)
return yearslayout
def plot_waveform_bokeh(filename,waveform_list,metadata_list,station_lat_list,\
station_lon_list, event_lat, event_lon, boundary_data, style_parameter):
xlabel_fontsize = style_parameter['xlabel_fontsize']
#
map_station_location_bokeh = ColumnDataSource(data=dict(map_lat_list=station_lat_list,\
map_lon_list=station_lon_list))
dot_default_index = 0
selected_dot_on_map_bokeh = ColumnDataSource(data=dict(lat=[station_lat_list[dot_default_index]],\
lon=[station_lon_list[dot_default_index]],\
index=[dot_default_index]))
map_view = Figure(plot_width=style_parameter['map_view_plot_width'], \
plot_height=style_parameter['map_view_plot_height'], \
y_range=[style_parameter['map_view_lat_min'],\
style_parameter['map_view_lat_max']], x_range=[style_parameter['map_view_lon_min'],\
style_parameter['map_view_lon_max']], tools=style_parameter['map_view_tools'],\
title=style_parameter['map_view_title'])
# ------------------------------
# add boundaries to map view
# country boundaries
map_view.multi_line(boundary_data['country']['longitude'],\
boundary_data['country']['latitude'],color='gray',\
line_width=2, level='underlay', nonselection_line_alpha=1.0,\
nonselection_line_color='gray')
# marine boundaries
map_view.multi_line(boundary_data['marine']['longitude'],\
boundary_data['marine']['latitude'],color='gray',\
level='underlay', nonselection_line_alpha=1.0,\
nonselection_line_color='gray')
# shoreline boundaries
map_view.multi_line(boundary_data['shoreline']['longitude'],\
boundary_data['shoreline']['latitude'],color='gray',\
line_width=2, nonselection_line_alpha=1.0, level='underlay',
nonselection_line_color='gray')
# state boundaries
map_view.multi_line(boundary_data['state']['longitude'],\
boundary_data['state']['latitude'],color='gray',\
level='underlay', nonselection_line_alpha=1.0,\
nonselection_line_color='gray')
#
map_view.triangle('map_lon_list', 'map_lat_list', source=map_station_location_bokeh, \
line_color='gray', size=style_parameter['marker_size'], fill_color='black',\
selection_color='black', selection_line_color='gray',\
selection_fill_alpha=1.0,\
nonselection_fill_alpha=1.0, nonselection_fill_color='black',\
nonselection_line_color='gray', nonselection_line_alpha=1.0)
map_view.triangle('lon','lat', source=selected_dot_on_map_bokeh,\
size=style_parameter['selected_marker_size'], line_color='black',fill_color='red')
map_view.asterisk([event_lon], [event_lat], size=style_parameter['event_marker_size'], line_width=3, line_color='red', \
fill_color='red')
# change style
map_view.title.text_font_size = style_parameter['title_font_size']
map_view.title.align = 'center'
map_view.title.text_font_style = 'normal'
map_view.xaxis.axis_label = style_parameter['map_view_xlabel']
map_view.xaxis.axis_label_text_font_style = 'normal'
map_view.xaxis.axis_label_text_font_size = xlabel_fontsize
map_view.xaxis.major_label_text_font_size = xlabel_fontsize
map_view.yaxis.axis_label = style_parameter['map_view_ylabel']
map_view.yaxis.axis_label_text_font_style = 'normal'
map_view.yaxis.axis_label_text_font_size = xlabel_fontsize
map_view.yaxis.major_label_text_font_size = xlabel_fontsize
map_view.xgrid.grid_line_color = None
map_view.ygrid.grid_line_color = None
map_view.toolbar.logo = None
map_view.toolbar_location = 'above'
map_view.toolbar_sticky = False
# --------------------------------------------------------
max_waveform_length = 0
max_waveform_amp = 0
ncurve = len(waveform_list)
for a_sta in waveform_list:
for a_trace in a_sta:
if len(a_trace) > max_waveform_length:
max_waveform_length = len(a_trace)
if np.max(np.abs(a_trace)) > max_waveform_amp:
max_waveform_amp = np.max(np.abs(a_trace))
#
plotting_list = []
for a_sta in waveform_list:
temp = []
for a_trace in a_sta:
if len(a_trace) < max_waveform_length:
a_trace = np.append(a_trace,np.zeros([(max_waveform_length-len(a_trace)),1]))
temp.append(list(a_trace))
plotting_list.append(temp)
#
time_list = []
for ista in range(len(plotting_list)):
a_sta = plotting_list[ista]
temp = []
for itr in range(len(a_sta)):
a_trace = a_sta[itr]
delta = metadata_list[ista][itr]['delta']
time = list(np.arange(len(a_trace))*delta)
temp.append(time)
#
time_list.append(temp)
#
reftime_label_list = []
channel_label_list = []
for ista in range(len(metadata_list)):
temp_ref = []
temp_channel = []
a_sta = metadata_list[ista]
for a_trace in a_sta:
temp_ref.append('Starting from '+a_trace['starttime'])
temp_channel.append(a_trace['network']+'_'+a_trace['station']+'_'+a_trace['channel'])
reftime_label_list.append(temp_ref)
channel_label_list.append(temp_channel)
# --------------------------------------------------------
curve_fig01 = Figure(plot_width=style_parameter['curve_plot_width'], plot_height=style_parameter['curve_plot_height'], \
y_range=(-max_waveform_amp*1.05,max_waveform_amp*1.05), \
x_range=(0,max_waveform_length),\
tools=['save','box_zoom','ywheel_zoom','xwheel_zoom','reset','crosshair','pan'])
#
curve_index = 0
select_curve_data = plotting_list[dot_default_index][curve_index]
select_curve_time = time_list[dot_default_index][curve_index]
selected_curve_data_bokeh01 = ColumnDataSource(data=dict(time=select_curve_time,amp=select_curve_data))
select_reftime_label = reftime_label_list[dot_default_index][curve_index]
selected_reftime_label_bokeh01 = ColumnDataSource(data=dict(x=[style_parameter['curve_reftime_label_x']],\
y=[style_parameter['curve_reftime_label_y']],\
label=[select_reftime_label]))
select_channel_label = channel_label_list[dot_default_index][curve_index]
selected_channel_label_bokeh01 = ColumnDataSource(data=dict(x=[style_parameter['curve_channel_label_x']],\
y=[style_parameter['curve_channel_label_y']],\
label=[select_channel_label]))
all_curve_data_bokeh = ColumnDataSource(data=dict(t=time_list, amp=plotting_list))
all_reftime_label_bokeh = ColumnDataSource(data=dict(label=reftime_label_list))
all_channel_label_bokeh = ColumnDataSource(data=dict(label=channel_label_list))
# plot waveform
curve_fig01.line('time','amp', source=selected_curve_data_bokeh01,\
line_color='black')
# add refference time as a label
curve_fig01.text('x', 'y', 'label', source=selected_reftime_label_bokeh01)
# add channel label
curve_fig01.text('x', 'y', 'label', source=selected_channel_label_bokeh01)
# change style
curve_fig01.title.text_font_size = style_parameter['title_font_size']
curve_fig01.title.align = 'center'
curve_fig01.title.text_font_style = 'normal'
curve_fig01.xaxis.axis_label = style_parameter['curve_xlabel']
curve_fig01.xaxis.axis_label_text_font_style = 'normal'
curve_fig01.xaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig01.xaxis.major_label_text_font_size = xlabel_fontsize
curve_fig01.yaxis.axis_label = style_parameter['curve_ylabel']
curve_fig01.yaxis.axis_label_text_font_style = 'normal'
curve_fig01.yaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig01.yaxis.major_label_text_font_size = xlabel_fontsize
curve_fig01.toolbar.logo = None
curve_fig01.toolbar_location = 'above'
curve_fig01.toolbar_sticky = False
# --------------------------------------------------------
curve_fig02 = Figure(plot_width=style_parameter['curve_plot_width'], plot_height=style_parameter['curve_plot_height'], \
y_range=(-max_waveform_amp*1.05,max_waveform_amp*1.05), \
x_range=(0,max_waveform_length),\
tools=['save','box_zoom','ywheel_zoom','xwheel_zoom','reset','crosshair','pan'])
#
curve_index = 1
select_curve_data = plotting_list[dot_default_index][curve_index]
select_curve_time = time_list[dot_default_index][curve_index]
selected_curve_data_bokeh02 = ColumnDataSource(data=dict(time=select_curve_time,amp=select_curve_data))
select_channel_label = channel_label_list[dot_default_index][curve_index]
selected_channel_label_bokeh02 = ColumnDataSource(data=dict(x=[style_parameter['curve_channel_label_x']],\
y=[style_parameter['curve_channel_label_y']],\
label=[select_channel_label]))
# plot waveform
curve_fig02.line('time','amp', source=selected_curve_data_bokeh02,\
line_color='black')
# add channel label
curve_fig02.text('x', 'y', 'label', source=selected_channel_label_bokeh02)
# change style
curve_fig02.title.text_font_size = style_parameter['title_font_size']
curve_fig02.title.align = 'center'
curve_fig02.title.text_font_style = 'normal'
curve_fig02.xaxis.axis_label = style_parameter['curve_xlabel']
curve_fig02.xaxis.axis_label_text_font_style = 'normal'
curve_fig02.xaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig02.xaxis.major_label_text_font_size = xlabel_fontsize
curve_fig02.yaxis.axis_label = style_parameter['curve_ylabel']
curve_fig02.yaxis.axis_label_text_font_style = 'normal'
curve_fig02.yaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig02.yaxis.major_label_text_font_size = xlabel_fontsize
curve_fig02.toolbar.logo = None
curve_fig02.toolbar_location = 'above'
curve_fig02.toolbar_sticky = False
# --------------------------------------------------------
curve_fig03 = Figure(plot_width=style_parameter['curve_plot_width'], plot_height=style_parameter['curve_plot_height'], \
y_range=(-max_waveform_amp*1.05,max_waveform_amp*1.05), \
x_range=(0,max_waveform_length),\
tools=['save','box_zoom','ywheel_zoom','xwheel_zoom','reset','crosshair','pan'])
#
curve_index = 2
select_curve_data = plotting_list[dot_default_index][curve_index]
select_curve_time = time_list[dot_default_index][curve_index]
selected_curve_data_bokeh03 = ColumnDataSource(data=dict(time=select_curve_time,amp=select_curve_data))
select_channel_label = channel_label_list[dot_default_index][curve_index]
selected_channel_label_bokeh03 = ColumnDataSource(data=dict(x=[style_parameter['curve_channel_label_x']],\
y=[style_parameter['curve_channel_label_y']],\
label=[select_channel_label]))
# plot waveform
curve_fig03.line('time','amp', source=selected_curve_data_bokeh03,\
line_color='black')
# add channel label
curve_fig03.text('x', 'y', 'label', source=selected_channel_label_bokeh03)
# change style
curve_fig03.title.text_font_size = style_parameter['title_font_size']
curve_fig03.title.align = 'center'
curve_fig03.title.text_font_style = 'normal'
curve_fig03.xaxis.axis_label = style_parameter['curve_xlabel']
curve_fig03.xaxis.axis_label_text_font_style = 'normal'
curve_fig03.xaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig03.xaxis.major_label_text_font_size = xlabel_fontsize
curve_fig03.yaxis.axis_label = style_parameter['curve_ylabel']
curve_fig03.yaxis.axis_label_text_font_style = 'normal'
curve_fig03.yaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig03.yaxis.major_label_text_font_size = xlabel_fontsize
curve_fig03.toolbar.logo = None
curve_fig03.toolbar_location = 'above'
curve_fig03.toolbar_sticky = False
# --------------------------------------------------------
map_station_location_bokeh.callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
map_station_location_bokeh=map_station_location_bokeh,\
selected_curve_data_bokeh01=selected_curve_data_bokeh01,\
selected_curve_data_bokeh02=selected_curve_data_bokeh02,\
selected_curve_data_bokeh03=selected_curve_data_bokeh03,\
selected_channel_label_bokeh01=selected_channel_label_bokeh01,\
selected_channel_label_bokeh02=selected_channel_label_bokeh02,\
selected_channel_label_bokeh03=selected_channel_label_bokeh03,\
selected_reftime_label_bokeh01=selected_reftime_label_bokeh01,\
all_reftime_label_bokeh=all_reftime_label_bokeh,\
all_channel_label_bokeh=all_channel_label_bokeh,\
all_curve_data_bokeh=all_curve_data_bokeh), code="""
var inds = Math.round(cb_obj.selected['1d'].indices)
selected_dot_on_map_bokeh.data['index'] = [inds]
var new_loc = map_station_location_bokeh.data
selected_dot_on_map_bokeh.data['lat'] = [new_loc['map_lat_list'][inds]]
selected_dot_on_map_bokeh.data['lon'] = [new_loc['map_lon_list'][inds]]
selected_dot_on_map_bokeh.change.emit()
selected_curve_data_bokeh01.data['t'] = all_curve_data_bokeh.data['t'][inds][0]
selected_curve_data_bokeh01.data['amp'] = all_curve_data_bokeh.data['amp'][inds][0]
selected_curve_data_bokeh01.change.emit()
selected_curve_data_bokeh02.data['t'] = all_curve_data_bokeh.data['t'][inds][1]
selected_curve_data_bokeh02.data['amp'] = all_curve_data_bokeh.data['amp'][inds][1]
selected_curve_data_bokeh02.change.emit()
selected_curve_data_bokeh03.data['t'] = all_curve_data_bokeh.data['t'][inds][2]
selected_curve_data_bokeh03.data['amp'] = all_curve_data_bokeh.data['amp'][inds][2]
selected_curve_data_bokeh03.change.emit()
selected_reftime_label_bokeh01.data['label'] = [all_reftime_label_bokeh.data['label'][inds][0]]
selected_reftime_label_bokeh01.change.emit()
selected_channel_label_bokeh01.data['label'] = [all_channel_label_bokeh.data['label'][inds][0]]
selected_channel_label_bokeh01.change.emit()
selected_channel_label_bokeh02.data['label'] = [all_channel_label_bokeh.data['label'][inds][1]]
selected_channel_label_bokeh02.change.emit()
selected_channel_label_bokeh03.data['label'] = [all_channel_label_bokeh.data['label'][inds][2]]
selected_channel_label_bokeh03.change.emit()
""")
curve_slider_callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
map_station_location_bokeh=map_station_location_bokeh,\
selected_curve_data_bokeh01=selected_curve_data_bokeh01,\
selected_curve_data_bokeh02=selected_curve_data_bokeh02,\
selected_curve_data_bokeh03=selected_curve_data_bokeh03,\
selected_channel_label_bokeh01=selected_channel_label_bokeh01,\
selected_channel_label_bokeh02=selected_channel_label_bokeh02,\
selected_channel_label_bokeh03=selected_channel_label_bokeh03,\
selected_reftime_label_bokeh01=selected_reftime_label_bokeh01,\
all_reftime_label_bokeh=all_reftime_label_bokeh,\
all_channel_label_bokeh=all_channel_label_bokeh,\
all_curve_data_bokeh=all_curve_data_bokeh),code="""
var inds = Math.round(cb_obj.value)
selected_dot_on_map_bokeh.data['index'] = [inds]
var new_loc = map_station_location_bokeh.data
selected_dot_on_map_bokeh.data['lat'] = [new_loc['map_lat_list'][inds]]
selected_dot_on_map_bokeh.data['lon'] = [new_loc['map_lon_list'][inds]]
selected_dot_on_map_bokeh.change.emit()
selected_curve_data_bokeh01.data['t'] = all_curve_data_bokeh.data['t'][inds][0]
selected_curve_data_bokeh01.data['amp'] = all_curve_data_bokeh.data['amp'][inds][0]
selected_curve_data_bokeh01.change.emit()
selected_curve_data_bokeh02.data['t'] = all_curve_data_bokeh.data['t'][inds][1]
selected_curve_data_bokeh02.data['amp'] = all_curve_data_bokeh.data['amp'][inds][1]
selected_curve_data_bokeh02.change.emit()
selected_curve_data_bokeh03.data['t'] = all_curve_data_bokeh.data['t'][inds][2]
selected_curve_data_bokeh03.data['amp'] = all_curve_data_bokeh.data['amp'][inds][2]
selected_curve_data_bokeh03.change.emit()
selected_reftime_label_bokeh01.data['label'] = [all_reftime_label_bokeh.data['label'][inds][0]]
selected_reftime_label_bokeh01.change.emit()
selected_channel_label_bokeh01.data['label'] = [all_channel_label_bokeh.data['label'][inds][0]]
selected_channel_label_bokeh01.change.emit()
selected_channel_label_bokeh02.data['label'] = [all_channel_label_bokeh.data['label'][inds][1]]
selected_channel_label_bokeh02.change.emit()
selected_channel_label_bokeh03.data['label'] = [all_channel_label_bokeh.data['label'][inds][2]]
selected_channel_label_bokeh03.change.emit()
""")
curve_slider = Slider(start=0, end=ncurve-1, value=style_parameter['curve_default_index'], \
step=1, title=style_parameter['curve_slider_title'], width=style_parameter['map_view_plot_width'],\
height=50, callback=curve_slider_callback)
# ==============================
# annotating text
annotating_fig01 = Div(text=style_parameter['annotating_html01'], \
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
annotating_fig02 = Div(text=style_parameter['annotating_html02'],\
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
# ==============================
output_file(filename,title=style_parameter['html_title'],mode=style_parameter['library_source'])
#
left_fig = Column(curve_slider, map_view, annotating_fig01, width=style_parameter['left_column_width'] )
right_fig = Column(curve_fig01, curve_fig02, curve_fig03, annotating_fig02, width=style_parameter['right_column_width'])
layout = Row(left_fig, right_fig)
save(layout)
def plot_dispersion_bokeh(filename, period_array, curve_data_array, boundary_data, style_parameter):
'''
Plot dispersion maps and curves using bokeh
Input:
filename is the filename of the resulting html file
period_array is a list of period
curve_data_array is a list of dispersion curves
boundary_data is a list of boundaries
style_parameter contains plotting parameters
Output:
None
'''
xlabel_fontsize = style_parameter['xlabel_fontsize']
# ==============================
# prepare data
map_data_all_slices_velocity = []
map_data_all_slices_period = []
map_data_all_slices_color = []
colorbar_data_all_left = []
colorbar_data_all_right = []
nperiod = len(period_array)
ncurve = len(curve_data_array)
ncolor = len(palette)
palette_r = palette[::-1]
colorbar_top = [0.1 for i in range(ncolor)]
colorbar_bottom = [0 for i in range(ncolor)]
for iperiod in range(nperiod):
one_slice_lat_list = []
one_slice_lon_list = []
one_slice_vel_list = []
map_period = period_array[iperiod]
for icurve in range(ncurve):
acurve = curve_data_array[icurve]
curve_lat = acurve['latitude']
curve_lon = acurve['longitude']
curve_vel = acurve['velocity']
curve_period = acurve['period']
one_slice_lat_list.append(curve_lat)
one_slice_lon_list.append(curve_lon)
if map_period in curve_period:
curve_period_index = curve_period.index(map_period)
one_slice_vel_list.append(curve_vel[curve_period_index])
else:
one_slice_vel_list.append(style_parameter['nan_value'])
# get color for dispersion values
one_slice_vel_mean = np.nanmean(one_slice_vel_list)
one_slice_vel_std = np.nanstd(one_slice_vel_list)
color_min = one_slice_vel_mean - one_slice_vel_std * style_parameter['spread_factor']
color_max = one_slice_vel_mean + one_slice_vel_std * style_parameter['spread_factor']
color_step = (color_max - color_min)*1./ncolor
one_slice_color_list = get_color_list(one_slice_vel_list,color_min,color_max,palette_r,\
style_parameter['nan_value'],style_parameter['nan_color'])
colorbar_left = np.linspace(color_min,color_max-color_step,ncolor)
colorbar_right = np.linspace(color_min+color_step,color_max,ncolor)
if one_slice_lat_list:
map_data_all_slices_velocity.append(one_slice_vel_list)
map_data_all_slices_period.append('Period: {0:6.1f} s'.format(map_period))
map_data_all_slices_color.append(one_slice_color_list)
colorbar_data_all_left.append(colorbar_left)
colorbar_data_all_right.append(colorbar_right)
# get location for all points
map_lat_list, map_lon_list = [], []
map_lat_label_list, map_lon_label_list = [], []
for i in range(ncurve):
acurve = curve_data_array[i]
map_lat_list.append(acurve['latitude'])
map_lon_list.append(acurve['longitude'])
map_lat_label_list.append('Lat: {0:12.3f}'.format(acurve['latitude']))
map_lon_label_list.append('Lon: {0:12.3f}'.format(acurve['longitude']))
# data for the map view plot
map_view_label_lon = style_parameter['map_view_period_label_lon']
map_view_label_lat = style_parameter['map_view_period_label_lat']
map_data_one_slice = map_data_all_slices_color[style_parameter['map_view_default_index']]
map_data_one_slice_period = map_data_all_slices_period[style_parameter['map_view_default_index']]
map_data_one_slice_bokeh = ColumnDataSource(data=dict(map_lat_list=map_lat_list,\
map_lon_list=map_lon_list,\
map_data_one_slice=map_data_one_slice))
map_data_one_slice_period_bokeh = ColumnDataSource(data=dict(lat=[map_view_label_lat], lon=[map_view_label_lon],
map_period=[map_data_one_slice_period]))
map_data_all_slices_bokeh = ColumnDataSource(data=dict(map_data_all_slices_color=map_data_all_slices_color,\
map_data_all_slices_period=map_data_all_slices_period))
# data for the colorbar
colorbar_data_one_slice = {}
colorbar_data_one_slice['colorbar_left'] = colorbar_data_all_left[style_parameter['map_view_default_index']]
colorbar_data_one_slice['colorbar_right'] = colorbar_data_all_right[style_parameter['map_view_default_index']]
colorbar_data_one_slice_bokeh = ColumnDataSource(data=dict(colorbar_top=colorbar_top,colorbar_bottom=colorbar_bottom,
colorbar_left=colorbar_data_one_slice['colorbar_left'],\
colorbar_right=colorbar_data_one_slice['colorbar_right'],\
palette_r=palette_r))
colorbar_data_all_slices_bokeh = ColumnDataSource(data=dict(colorbar_data_all_left=colorbar_data_all_left,\
colorbar_data_all_right=colorbar_data_all_right))
# data for dispersion curves
curve_default_index = style_parameter['curve_default_index']
selected_dot_on_map_bokeh = ColumnDataSource(data=dict(lat=[map_lat_list[curve_default_index]],\
lon=[map_lon_list[curve_default_index]],\
color=[map_data_one_slice[curve_default_index]],\
index=[curve_default_index]))
selected_curve_data = curve_data_array[curve_default_index]
selected_curve_data_bokeh = ColumnDataSource(data=dict(curve_period=selected_curve_data['period'],\
curve_velocity=selected_curve_data['velocity']))
period_all = []
velocity_all = []
for acurve in curve_data_array:
period_all.append(acurve['period'])
velocity_all.append(acurve['velocity'])
curve_data_all_bokeh = ColumnDataSource(data=dict(period_all=period_all, velocity_all=velocity_all))
selected_curve_lat_label_bokeh = ColumnDataSource(data=dict(x=[style_parameter['curve_lat_label_x']], \
y=[style_parameter['curve_lat_label_y']],\
lat_label=[map_lat_label_list[curve_default_index]]))
selected_curve_lon_label_bokeh = ColumnDataSource(data=dict(x=[style_parameter['curve_lon_label_x']], \
y=[style_parameter['curve_lon_label_y']],\
lon_label=[map_lon_label_list[curve_default_index]]))
all_curve_lat_label_bokeh = ColumnDataSource(data=dict(map_lat_label_list=map_lat_label_list))
all_curve_lon_label_bokeh = ColumnDataSource(data=dict(map_lon_label_list=map_lon_label_list))
# ==============================
map_view = Figure(plot_width=style_parameter['map_view_plot_width'], \
plot_height=style_parameter['map_view_plot_height'], \
y_range=[style_parameter['map_view_lat_min'],\
style_parameter['map_view_lat_max']], x_range=[style_parameter['map_view_lon_min'],\
style_parameter['map_view_lon_max']], tools=style_parameter['map_view_tools'],\
title=style_parameter['map_view_title'])
# ------------------------------
# add boundaries to map view
# country boundaries
map_view.multi_line(boundary_data['country']['longitude'],\
boundary_data['country']['latitude'],color='black',\
line_width=2, level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# marine boundaries
map_view.multi_line(boundary_data['marine']['longitude'],\
boundary_data['marine']['latitude'],color='black',\
level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# shoreline boundaries
map_view.multi_line(boundary_data['shoreline']['longitude'],\
boundary_data['shoreline']['latitude'],color='black',\
line_width=2, level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# state boundaries
map_view.multi_line(boundary_data['state']['longitude'],\
boundary_data['state']['latitude'],color='black',\
level='underlay',nonselection_line_alpha=1.0,\
nonselection_line_color='black')
# ------------------------------
# add period label
map_view.rect(style_parameter['map_view_period_box_lon'], style_parameter['map_view_period_box_lat'], \
width=style_parameter['map_view_period_box_width'], height=style_parameter['map_view_period_box_height'], \
width_units='screen',height_units='screen', color='#FFFFFF', line_width=1., line_color='black', level='underlay')
map_view.text('lon', 'lat', 'map_period', source=map_data_one_slice_period_bokeh,\
text_font_size=style_parameter['annotating_text_font_size'],text_align='left',level='underlay')
# ------------------------------
# plot dots
map_view.circle('map_lon_list', 'map_lat_list', color='map_data_one_slice', \
source=map_data_one_slice_bokeh, size=style_parameter['marker_size'],\
line_width=0.2, line_color='black', alpha=1.0,\
selection_color='map_data_one_slice', selection_line_color='black',\
selection_fill_alpha=1.0,\
nonselection_fill_alpha=1.0, nonselection_fill_color='map_data_one_slice',\
nonselection_line_color='black', nonselection_line_alpha=1.0)
map_view.circle('lon', 'lat', color='color', source=selected_dot_on_map_bokeh, \
line_color='#00ff00', line_width=4.0, alpha=1.0, \
size=style_parameter['selected_marker_size'])
# ------------------------------
# change style
map_view.title.text_font_size = style_parameter['title_font_size']
map_view.title.align = 'center'
map_view.title.text_font_style = 'normal'
map_view.xaxis.axis_label = style_parameter['map_view_xlabel']
map_view.xaxis.axis_label_text_font_style = 'normal'
map_view.xaxis.axis_label_text_font_size = xlabel_fontsize
map_view.xaxis.major_label_text_font_size = xlabel_fontsize
map_view.yaxis.axis_label = style_parameter['map_view_ylabel']
map_view.yaxis.axis_label_text_font_style = 'normal'
map_view.yaxis.axis_label_text_font_size = xlabel_fontsize
map_view.yaxis.major_label_text_font_size = xlabel_fontsize
map_view.xgrid.grid_line_color = None
map_view.ygrid.grid_line_color = None
map_view.toolbar.logo = None
map_view.toolbar_location = 'above'
map_view.toolbar_sticky = False
# ==============================
# plot colorbar
colorbar_fig = Figure(tools=[], y_range=(0,0.1),plot_width=style_parameter['map_view_plot_width'], \
plot_height=style_parameter['colorbar_plot_height'],title=style_parameter['colorbar_title'])
colorbar_fig.toolbar_location=None
colorbar_fig.quad(top='colorbar_top',bottom='colorbar_bottom',left='colorbar_left',right='colorbar_right',\
fill_color='palette_r',source=colorbar_data_one_slice_bokeh)
colorbar_fig.yaxis[0].ticker=FixedTicker(ticks=[])
colorbar_fig.xgrid.grid_line_color = None
colorbar_fig.ygrid.grid_line_color = None
colorbar_fig.xaxis.axis_label_text_font_size = xlabel_fontsize
colorbar_fig.xaxis.major_label_text_font_size = xlabel_fontsize
colorbar_fig.xaxis[0].formatter = PrintfTickFormatter(format="%5.2f")
colorbar_fig.title.text_font_size = xlabel_fontsize
colorbar_fig.title.align = 'center'
colorbar_fig.title.text_font_style = 'normal'
# ==============================
curve_fig = Figure(plot_width=style_parameter['curve_plot_width'], plot_height=style_parameter['curve_plot_height'], \
y_range=(style_parameter['curve_y_min'],style_parameter['curve_y_max']), \
x_range=(style_parameter['curve_x_min'],style_parameter['curve_x_max']),x_axis_type='log',\
tools=['save','box_zoom','wheel_zoom','reset','crosshair','pan'],
title=style_parameter['curve_title'])
# ------------------------------
curve_fig.rect([style_parameter['curve_label_box_x']], [style_parameter['curve_label_box_y']], \
width=style_parameter['curve_label_box_width'], height=style_parameter['curve_label_box_height'], \
width_units='screen', height_units='screen', color='#FFFFFF', line_width=1., line_color='black', level='underlay')
curve_fig.text('x', 'y', \
'lat_label', source=selected_curve_lat_label_bokeh)
curve_fig.text('x', 'y', \
'lon_label', source=selected_curve_lon_label_bokeh)
# ------------------------------
curve_fig.line('curve_period', 'curve_velocity', source=selected_curve_data_bokeh, color='black')
curve_fig.circle('curve_period', 'curve_velocity', source=selected_curve_data_bokeh, size=5, color='black')
# ------------------------------
curve_fig.title.text_font_size = style_parameter['title_font_size']
curve_fig.title.align = 'center'
curve_fig.title.text_font_style = 'normal'
curve_fig.xaxis.axis_label = style_parameter['curve_xlabel']
curve_fig.xaxis.axis_label_text_font_style = 'normal'
curve_fig.xaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig.xaxis.major_label_text_font_size = xlabel_fontsize
curve_fig.yaxis.axis_label = style_parameter['curve_ylabel']
curve_fig.yaxis.axis_label_text_font_style = 'normal'
curve_fig.yaxis.axis_label_text_font_size = xlabel_fontsize
curve_fig.yaxis.major_label_text_font_size = xlabel_fontsize
curve_fig.xgrid.grid_line_dash = [4, 2]
curve_fig.ygrid.grid_line_dash = [4, 2]
curve_fig.xaxis[0].formatter = PrintfTickFormatter(format="%4.0f")
curve_fig.toolbar.logo = None
curve_fig.toolbar_location = 'above'
curve_fig.toolbar_sticky = False
# ==============================
map_data_one_slice_bokeh.callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
map_data_one_slice_bokeh=map_data_one_slice_bokeh,\
selected_curve_data_bokeh=selected_curve_data_bokeh,\
curve_data_all_bokeh=curve_data_all_bokeh,\
selected_curve_lat_label_bokeh=selected_curve_lat_label_bokeh,\
selected_curve_lon_label_bokeh=selected_curve_lon_label_bokeh,\
all_curve_lat_label_bokeh=all_curve_lat_label_bokeh,\
all_curve_lon_label_bokeh=all_curve_lon_label_bokeh), code="""
var inds = Math.round(cb_obj.selected['1d'].indices)
selected_dot_on_map_bokeh.data['index'] = [inds]
var new_slice = map_data_one_slice_bokeh.data
selected_dot_on_map_bokeh.data['lat'] = [new_slice['map_lat_list'][inds]]
selected_dot_on_map_bokeh.data['lon'] = [new_slice['map_lon_list'][inds]]
selected_dot_on_map_bokeh.data['color'] = [new_slice['map_data_one_slice'][inds]]
selected_dot_on_map_bokeh.change.emit()
selected_curve_data_bokeh.data['curve_period'] = curve_data_all_bokeh.data['period_all'][inds]
selected_curve_data_bokeh.data['curve_velocity'] = curve_data_all_bokeh.data['velocity_all'][inds]
selected_curve_data_bokeh.change.emit()
var all_lat_labels = all_curve_lat_label_bokeh.data['map_lat_label_list']
var all_lon_labels = all_curve_lon_label_bokeh.data['map_lon_label_list']
selected_curve_lat_label_bokeh.data['lat_label'] = [all_lat_labels[inds]]
selected_curve_lon_label_bokeh.data['lon_label'] = [all_lon_labels[inds]]
selected_curve_lat_label_bokeh.change.emit()
selected_curve_lon_label_bokeh.change.emit()
""")
# ==============================
period_slider_callback = CustomJS(args=dict(map_data_all_slices_bokeh=map_data_all_slices_bokeh,\
map_data_one_slice_bokeh=map_data_one_slice_bokeh,\
colorbar_data_all_slices_bokeh=colorbar_data_all_slices_bokeh, \
colorbar_data_one_slice_bokeh=colorbar_data_one_slice_bokeh,\
selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
map_data_one_slice_period_bokeh=map_data_one_slice_period_bokeh),\
code="""
var p_index = Math.round(cb_obj.value)
var map_data_all_slices = map_data_all_slices_bokeh.data
var map_data_new_slice = map_data_all_slices['map_data_all_slices_color'][p_index]
map_data_one_slice_bokeh.data['map_data_one_slice'] = map_data_new_slice
map_data_one_slice_bokeh.change.emit()
var color_data_all_slices = colorbar_data_all_slices_bokeh.data
colorbar_data_one_slice_bokeh.data['colorbar_left'] = color_data_all_slices['colorbar_data_all_left'][p_index]
colorbar_data_one_slice_bokeh.data['colorbar_right'] = color_data_all_slices['colorbar_data_all_right'][p_index]
colorbar_data_one_slice_bokeh.change.emit()
var selected_index = selected_dot_on_map_bokeh.data['index']
selected_dot_on_map_bokeh.data['color'] = [map_data_new_slice[selected_index]]
selected_dot_on_map_bokeh.change.emit()
map_data_one_slice_period_bokeh.data['map_period'] = [map_data_all_slices['map_data_all_slices_period'][p_index]]
map_data_one_slice_period_bokeh.change.emit()
""")
period_slider = Slider(start=0, end=nperiod-1, value=style_parameter['map_view_default_index'], \
step=1, title=style_parameter['period_slider_title'], \
width=style_parameter['period_slider_plot_width'],\
height=50, callback=period_slider_callback)
# ==============================
curve_slider_callback = CustomJS(args=dict(selected_dot_on_map_bokeh=selected_dot_on_map_bokeh,\
map_data_one_slice_bokeh=map_data_one_slice_bokeh,\
selected_curve_data_bokeh=selected_curve_data_bokeh,\
curve_data_all_bokeh=curve_data_all_bokeh,\
selected_curve_lat_label_bokeh=selected_curve_lat_label_bokeh,\
selected_curve_lon_label_bokeh=selected_curve_lon_label_bokeh,\
all_curve_lat_label_bokeh=all_curve_lat_label_bokeh,\
all_curve_lon_label_bokeh=all_curve_lon_label_bokeh),\
code="""
var c_index = Math.round(cb_obj.value)
var one_slice = map_data_one_slice_bokeh.data
selected_dot_on_map_bokeh.data['index'] = [c_index]
selected_dot_on_map_bokeh.data['lat'] = [one_slice['map_lat_list'][c_index]]
selected_dot_on_map_bokeh.data['lon'] = [one_slice['map_lon_list'][c_index]]
selected_dot_on_map_bokeh.data['color'] = [one_slice['map_data_one_slice'][c_index]]
selected_dot_on_map_bokeh.change.emit()
selected_curve_data_bokeh.data['curve_period'] = curve_data_all_bokeh.data['period_all'][c_index]
selected_curve_data_bokeh.data['curve_velocity'] = curve_data_all_bokeh.data['velocity_all'][c_index]
selected_curve_data_bokeh.change.emit()
var all_lat_labels = all_curve_lat_label_bokeh.data['map_lat_label_list']
var all_lon_labels = all_curve_lon_label_bokeh.data['map_lon_label_list']
selected_curve_lat_label_bokeh.data['lat_label'] = [all_lat_labels[c_index]]
selected_curve_lon_label_bokeh.data['lon_label'] = [all_lon_labels[c_index]]
selected_curve_lat_label_bokeh.change.emit()
selected_curve_lon_label_bokeh.change.emit()
""")
curve_slider = Slider(start=0, end=ncurve-1, value=style_parameter['curve_default_index'], \
step=1, title=style_parameter['curve_slider_title'], width=style_parameter['curve_plot_width'],\
height=50, callback=curve_slider_callback)
# ==============================
# annotating text
annotating_fig01 = Div(text=style_parameter['annotating_html01'], \
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
annotating_fig02 = Div(text=style_parameter['annotating_html02'],\
width=style_parameter['annotation_plot_width'], height=style_parameter['annotation_plot_height'])
# ==============================
output_file(filename,title=style_parameter['html_title'],mode=style_parameter['library_source'])
left_fig = Column(period_slider, map_view, colorbar_fig, annotating_fig01,\
width=style_parameter['left_column_width'] )
right_fig = Column(curve_slider, curve_fig, annotating_fig02, \
width=style_parameter['right_column_width'] )
layout = Row(left_fig, right_fig)
save(layout)
def create_objects(cls, symbol, df, securities):
descr_box = Paragraph(text='content loading...')
btn_close_loading = Button(label='Close Loading')
dialog_loading = Dialog(title='loading',
content=vplot(descr_box),
name='loading_dialog',
buttons=[btn_close_loading],
visible=False)
source_data = dict(df)
main_source = ColumnDataSource(dict(df))
source = ColumnDataSource(source_data)
# TODO: REMOVE THIS COMMENTED CODE! IT'S JUST THE PREVIOUS
# VERSION USED BEFORE NEW P&D Cached results and algorithm
# get the cached results of the P&D algorithm computed with the
# "default" configuration
# intervals = utils.cached_pumps.get(symbol, pumps.to_dicts(((),(),(),(),(),())))
# intervals['bottom'] = [0] * len(intervals['start'])
# intervals['values'] = [max(df['price'])] * len(intervals['start'])
#
# intervals = pd.DataFrame(intervals)
# new version
stats = utils.get_symbols_cached_stats()[symbol]
intervals = pd.DataFrame(stats)
intervals['bottom'] = [0] * len(intervals['start'])
intervals['values'] = [max(df['price'])] * len(intervals['start'])
conv = lambda x: utils.to_seconds(pd.to_datetime(x))
intervals = intervals[
(pd.to_datetime(intervals['start']) > conv(config.date_range[0])) &
(pd.to_datetime(intervals['start']) < conv(config.date_range[1]))]
# Create P&Ds intervals DataSource
intervals_source = ColumnDataSource(intervals)
source.tags = ['main_source']
trends = utils.load_trends_data(symbol, start_date=min(df['dt']))
trends_source = ColumnDataSource(trends)
trades = Slider(title="trades",
name='trades',
value=0,
start=0,
end=124,
step=1)
# Selectors
symbol = Select.create(options=securities,
value=symbol,
name='symbol',
title="")
window_selector = Select.create(options=['---'],
name='period_selector',
title="Search intervals with:")
symbol_filter = Select.create(
options=['All', 'Stocks with Spam', 'Stocks without Spam'],
name='symbol_filter',
title="Filter Symbols:",
value='Stocks with Spam')
callback = Callback(args={
'symbol_filter': symbol_filter,
'dialog_loading': dialog_loading
},
code=callbacks.symbol_filter)
symbol_filter.callback = callback
btn_detect_pumps = Button(label='Configure P&D Detection',
name='config_pumps')
main_tab = Panel(title="Main")
tabs = Tabs()
# Create STOCKS TABLE
ranks = utils.get_pumps_rank()
# quotient_metrics = utils.get_quotient_metrics()
# ranks['quotient'] = quotient_metrics['quotient']
foo = lambda x: utils.spams_count.get(x, 0)
ranks['spams'] = map(foo, ranks['symbol'])
ranks = ranks.sort(['spams', 'vol_quotient'], ascending=False)
cls._pre_filtered_ranks = {
'All': {k: ranks[k]
for k in ranks.columns},
'Stocks with Spam':
dict(ranks[ranks['spams'] > 0].sort('vol_quotient',
ascending=False)),
'Stocks without Spam':
dict(ranks[ranks['spams'] == 0].sort('vol_quotient',
ascending=False)),
}
source_stocks_rank = ColumnDataSource(cls._pre_filtered_ranks['All'])
table_stocks_rank = DataTable(
source=source_stocks_rank,
width=560,
height=450,
selectable=True,
editable=True,
columns=[
TableColumn(field='symbol',
title='symbol',
width=130,
editor=StringEditor()),
TableColumn(field='vol_quotient',
title='volume ratio',
editor=StringEditor(),
default_sort='descending'),
TableColumn(field='risk_score',
title='risk',
width=100,
editor=StringEditor(),
default_sort='descending'),
TableColumn(field='spams',
title='spams',
width=130,
editor=StringEditor(),
default_sort='descending'),
])
callback = Callback(args={
'tr': table_stocks_rank,
'sr': source_stocks_rank,
'symb': symbol,
'dialog_loading': dialog_loading
},
code=callbacks.source_stocks_rank)
source_stocks_rank.callback = callback
return locals()
def paired_plots(data, title=None, text_cols=['seqname']):
"""Create an interactive scatterplot. ``data`` is a pandas dataframe
with (at least) column 'dist' in addition to columns
containing other features. ``text_cols`` is a list of columns to
include in the DataTable.
Returns a pair of plot objects (plt, tab).
"""
data['i'] = range(len(data))
data = data.fillna('')
data['select_alpha'] = data['is_type'].apply(lambda x: 0 if x else 1)
data['unselect_alpha'] = data['is_type'].apply(lambda x: 0 if x else 0.2)
data['out_alpha'] = data['is_out'].apply(lambda x: 1 if x else 0)
data['size'] = 10
data['triangle_size'] = data['size'] * 1.1
source = ColumnDataSource(data=data)
types = ColumnDataSource(data=data[data['is_type']])
# start with is_outs in table
table_rows = ColumnDataSource(data=data[data['is_out']])
tools = [
'box_select', 'lasso_select', 'resize', 'box_zoom', 'pan', 'reset',
'tap'
]
step_plt = figure(
title=title,
plot_width=600,
plot_height=600,
tools=tools,
x_axis_label='rank order',
y_axis_label='distance to centroid (%)',
)
# underlying markers are visible only when selected
step_plt.scatter(source=ColumnDataSource(data=data[data['is_out']]),
x='i',
y='dist_from_cent',
color='red',
fill_alpha='out_alpha',
line_alpha='select_alpha',
marker='circle',
size='size')
# underlying markers are visible only when selected
step_plt.scatter(source=table_rows,
fill_alpha=0,
line_alpha='select_alpha',
x='i',
y='dist_from_cent',
color='gray',
marker='circle',
size='size')
step_plt.scatter(source=types,
fill_alpha='out_alpha',
color='red',
line_alpha=1,
line_color='red',
x='i',
y='dist_from_cent',
marker='triangle',
size='triangle_size')
step_plt.scatter(source=source,
fill_alpha=0,
line_alpha='unselect_alpha',
line_color='gray',
x='i',
y='dist_from_cent',
marker='circle',
size='size')
pca_plt = figure(title=None,
plot_width=600,
plot_height=600,
tools=tools,
x_axis_label='distance',
y_axis_label='distance')
pca_plt.circle(
source=ColumnDataSource(data=data[data['is_out']]),
size='size',
fill_alpha='out_alpha',
line_alpha='select_alpha',
x='x',
y='y',
color='red',
)
# underlying markers are visible only when selected
pca_plt.circle(
source=table_rows,
line_alpha='select_alpha',
fill_alpha=0,
size='size',
x='x',
y='y',
color='gray',
)
pca_plt.triangle(
source=types,
line_color='red',
fill_alpha='out_alpha',
line_alpha=1,
size='triangle_size',
x='x',
y='y',
)
pca_plt.circle(
source=source,
fill_alpha=0,
line_alpha='unselect_alpha',
line_color='gray',
size='size',
x='x',
y='y',
)
formatters = {
'seqname':
bokeh.models.widgets.tables.HTMLTemplateFormatter(
template='<a href="https://www.ncbi.nlm.nih.gov/nuccore/'
'<%= version %>">'
'<div title="<%= value %>"><%= value %></div></a>'),
'type strain hit':
bokeh.models.widgets.tables.HTMLTemplateFormatter(),
'description':
bokeh.models.widgets.tables.HTMLTemplateFormatter(
template='<div title="<%= value %>"><%= value %></div>'),
'identity':
bokeh.models.widgets.tables.NumberFormatter(format='0.0%'),
'dist':
bokeh.models.widgets.tables.NumberFormatter(format='0.00%')
}
tab = DataTable(source=table_rows,
columns=[
TableColumn(field=col,
title=TABLE_HEADER.get(col, col),
width=TABLE_WIDTHS[col],
formatter=formatters.get(col, None))
for col in text_cols
],
fit_columns=False,
row_headers=False,
width=1200,
height=1200,
sortable=True)
code = """
var inds = cb_obj.selected['1d'].indices;
var data = source.data;
var text_data = text_source.data;
for (var column in text_data){
text_data[column] = [];
for (var i = 0; i < inds.length; i++) {
var ind = inds[i];
text_data[column].push(data[column][ind])
}
}
source.trigger('change');
text_source.trigger('change');
table.trigger('change');
"""
# note that the callback is added to the source for the scatter plot only
source.callback = CustomJS(args=dict(source=source,
table=tab,
text_source=table_rows),
code=code)
return step_plt, pca_plt, tab
def draw_bokeh(X, cluster, sentences):
"""Dimension reduction with PCA and plotting with bokeh """
X_reduced = PCA(n_components=2).fit_transform(X)
colors = [Category10[10][i % 10] for i in cluster.labels_]
sizes = np.ones(len(X)) * 10
sizes[cluster.cluster_centers_indices_] = 20
source = ColumnDataSource(data=dict(
eigenv1=[x[0] for x in X_reduced],
eigenv2=[x[1] for x in X_reduced],
s=sentences,
c=colors,
size=sizes,
))
hover = HoverTool(tooltips=[("index",
"$index"), ("(x,y)",
"@eigenv1, @eigenv2)"), ("s",
"@s")])
p = figure(title="reforma_educativa",
tools=[hover, ResetTool(), BoxSelectTool()])
p.toolbar.logo = None
p.xaxis.axis_label = '1st eigenv'
p.yaxis.axis_label = '2nd eigenv'
p.scatter('eigenv1',
'eigenv2',
size='size',
fill_color='c',
fill_alpha=1,
line_color='c',
nonselection_fill_color="c",
nonselection_fill_alpha=0.5,
source=source)
div = Div(width=600)
text_input = TextAreaInput(title="Text: ", rows=50, max_length=2000)
layout = row(p, div, text_input)
source.callback = CustomJS(args=dict(source=source,
div=div,
sentences=sentences,
colors=colors),
code="""
var inds = source.selected.indices
div.text = ""
for (var i = 0; i < sentences.length; i++) {
if (inds.includes(i)) {
div.text += '<p style="background-color:' + colors[i] + ';">' + sentences[i] + "</p>"
}
else {
r = parseInt(colors[i].substring(1, 3), 16)
g = parseInt(colors[i].substring(3, 5), 16)
b = parseInt(colors[i].substring(5, 7), 16)
a = 0.3
rgba = r + "," + g + "," + b + "," + a
div.text += '<p style="background-color:rgba(' + rgba + ');">' + sentences[i] + "</p>"
}
}
""")
output_file("reforma_educativa.html", title="Sentence embedding example")
show(layout)
