def plotting(*name):
try:
df = plot_df.copy()
except:
print 'run interval_rate() first'
return
from bokeh.io import output_file, output_notebook, show
from bokeh.charts import Line
from bokeh.layouts import row, column
p = Line(df[['day1', 'day7', 'matched']],
plot_width=950,
plot_height=400,
legend='top_right')
# q=Line(df[['matched']],plot_width=950,plot_height=400,color='blue')
r = Line(df[['single_day_retention']],
plot_width=950,
plot_height=400,
color='purple')
layout = column(p, r)
if len(name) > 0:
output_file(name[0] + '.html')
show(layout)
else:
output_notebook()
show(layout)
def fnCreate_Chart_Line(df):
pd.options.html.border = 1
plot = Line(df,
title="Date wise tweet graph",
legend=False,
xlabel='Date',
ylabel='Count') #, 'green', 'blue']
plot.logo = None
script, div = components(plot, CDN)
return script, div
def fnCreate_Chart_MultiLine(df):
pd.options.html.border = 1
plot = Line(df,
title="Likes vs retweets visualization",
legend="top_left",
xlabel='Date',
ylabel='Count')
plot.logo = None
script, div = components(plot, CDN)
return script, div
def createBokehChart(self):
keyFields = self.getKeyFields()
valueFields = self.getValueFields()
data = self.getWorkingPandasDataFrame().sort_values(keyFields[0])
subplots = self.options.get("lineChartType", "grouped") == "subplots"
clusterby = self.options.get("clusterby")
figs = []
if clusterby is None:
if subplots:
for valueField in valueFields:
figs.append(
Line(data,
x=keyFields[0],
y=valueField,
legend=self.showLegend(),
plot_width=int(800 / len(valueFields))))
else:
figs.append(
Line(data,
x=keyFields[0],
y=valueFields,
color=valueFields,
legend=self.showLegend()))
else:
if subplots:
self.addMessage(
"Warning: 'Cluster By' ignored when you have multiple Value Fields but subplots options selected"
)
for valueField in valueFields:
figs.append(
Line(data,
x=keyFields[0],
y=valueField,
legend=self.showLegend(),
plot_width=int(800 / len(valueFields))))
else:
if len(valueFields) > 1:
self.addMessage(
"Warning: 'Cluster By' ignored when you have multiple Value Fields but subplots option is not selected"
)
else:
self.addMessage(
"Warning: 'Cluster By' ignored when grouped option with multiple Value Fields is selected"
)
figs.append(
Line(data,
x=keyFields[0],
y=valueFields,
color=valueFields,
legend=self.showLegend()))
return figs
def fnCreate_Chart_MultiLine(df,strS,j=1):
pd.options.html.border=1
sL=False
if j ==1:
sL="top_center"
plot = Line(df, title="Average of Ticket Price Date wise" + strS, legend=sL, xlabel='Travel Date',ylabel='Average Ticket Price')
plot.legend.label_text_font_size = "7pt"
plot.legend.orientation = "horizontal"
plot.legend.click_policy="hide"
plot.logo=None
script, div = components(plot,CDN)
return script, div
def plot_state(state_data, varaible_names=[]):
# plot = figure(title='State Variables', x_axis_label='step', y_axis_label='level')
# data = []
data = dict()
for v in varaible_names:
data[v] = state_data[v]
# plot.line(range(len(data[v])), data[v], legend=v)
# xyvalues = np.array([[2, 3, 7, 5, 26], [12, 33, 47, 15, 126], [22, 43, 10, 25, 26]])
# xyvalues = np.array(data)
plot = Step(data,
title="state variables - step graph",
legend="top_left",
ylabel='',
palette=["red", "green", "blue", "orange"])
plot_line = Line(data,
title="state variables - line graph",
legend="top_left",
ylabel='',
palette=["red", "green", "blue", "orange"])
# output_file('line.html')
# show(plot)
return (plot, plot_line)
def abc_model(abc_id, model_id, t):
history = app.config["HISTORY"]
history.id = abc_id
if t == "max":
t = history.max_t
else:
t = int(t)
df, w = history.get_distribution(model_id, t)
df["CDF"] = w
tabs = []
model_ids = history.get_model_probabilities().columns
for parameter in [col for col in df if col != "CDF"]:
plot_df = df[["CDF", parameter]].sort_values(parameter)
plot_df_cumsum = plot_df.cumsum()
plot_df_cumsum[parameter] = plot_df[parameter]
p = Panel(child=Line(x=parameter, y="CDF", data=plot_df_cumsum),
title=parameter)
tabs.append(p)
if len(tabs) == 0:
plot = PlotScriptDiv("", "This model has no Parameters")
else:
plot = PlotScriptDiv(*components(Tabs(tabs=tabs)))
return render_template("model.html",
abc_id=abc_id,
model_id=model_id,
plot=plot,
BOKEH=BOKEH,
model_ids=model_ids,
t=t,
available_t=list(range(history.max_t + 1)))
def loss_accuracy_plot(df, x, y):
df = df.fillna(0)
plot = Row(*[Line(df, x, y, legend=True) for y in y])
script, div = components(plot, INLINE)
js_resources = RESOURCE.render_js()
css_resources = RESOURCE.render_css()
return Plot(js_resources, css_resources, script, div)
def line(self,
dataframe,
x=None,
y=None,
width=None,
height=None,
groups=None,
palette=None,
title="Line",
xaxis_label=None,
yaxis_label=None):
palette = self.__default_options__.get(
'palette', None) if palette is None else palette
width = self.__default_options__.get('width',
None) if width is None else width
width, height = self._width_height(width, height)
line = Line(dataframe,
x=x,
y=y,
color=groups,
width=width,
height=height,
palette=palette,
title=title,
legend=True)
if xaxis_label:
line._xaxis.axis_label = xaxis_label
if yaxis_label:
line._yaxis.axis_label = yaxis_label
return line
def line(self,
dataframe,
width=None,
height=None,
palette=None,
title="Line",
x_axis_label=None,
y_axis_label=None,
grid=None):
palette = self.get_option('palette') if palette is None else palette
width = self.get_option('width') if width is None else width
if not height:
width, height = self.golden_ratio(width, height)
palette = self._palette(palette, len(dataframe.index))
line = Line(dataframe.T,
legend="top_right",
color=palette,
ylabel=y_axis_label,
xlabel=y_axis_label,
title=title,
width=width,
height=height)
if grid is not None:
grid.append(line)
return grid
return line
def doPlot11(data, nrDataSource):
p = Line(data,
title="Line graph: " + nrDataSource['name'],
xlabel=data.columns[0],
ylabel=data.columns[1],
responsive=True)
c = components(p, resources=None, wrap_script=False, wrap_plot_info=True)
return c
def showBacktestingResult(self):
"""显示回测结果"""
d = self.calculateBacktestingResult()
# 输出
self.output('-' * 30)
self.output('First Trade：\t%s' % d['timeList'][0])
self.output('Last Trade：\t%s' % d['timeList'][-1])
self.output('Total Trades：\t%s' % formatNumber(d['totalResult']))
self.output('Total Return：\t%s' % formatNumber(d['capital']))
self.output('Maximum Drawdown: \t%s' %
formatNumber(min(d['drawdownList'])))
self.output('Ave Trade：\t%s' %
formatNumber(d['capital'] / d['totalResult']))
self.output('Ave Slippage：\t%s' %
formatNumber(d['totalSlippage'] / d['totalResult']))
self.output('Ave Commission：\t%s' %
formatNumber(d['totalCommission'] / d['totalResult']))
self.output('Win Ratio\t\t%s%%' % formatNumber(d['winningRate']))
self.output('Ave Win\t%s' % formatNumber(d['averageWinning']))
self.output('Ave Loss\t%s' % formatNumber(d['averageLosing']))
self.output('Profit Factor：\t%s' % formatNumber(d['profitLossRatio']))
# Use Bokeh to plot
from bokeh.charts import Area, Line, Histogram
from bokeh.layouts import column
from bokeh.io import show
plotdata = {
"TradeN": range(len(d['capitalList'])),
"Equity Curve": d['capitalList'],
"Maximum Drawdown": d['drawdownList'],
"Profit/Loss": d['pnlList']
}
f1 = Line(plotdata,
x="TradeN",
y="Equity Curve",
color="blue",
width=1000,
height=300)
f2 = Area(plotdata,
x="TradeN",
y="Maximum Drawdown",
color="tomato",
width=1000,
height=300)
f3 = Histogram(plotdata,
values="Profit/Loss",
bins=30,
color="green",
width=1000,
height=300)
show(column(f1, f2, f3))
def create_line(data):
""" Convenience function to create a new line chart with the right args """
return Line(data,
x='year',
y=countries,
legend=True,
width=1400,
height=300,
ylabel='Energy use per capita',
palette=['purple', 'green', 'blue', 'pink'])
def plot_close(tickr):
qd.ApiConfig.api_key = "FtpsTU3J-q2x2pa2KBqV"
tickrfinal=''.join(('WIKI/',tickr))
mydata=qd.get(tickrfinal,start_date="2017-5-1", end_date="2017-5-31")
print list(mydata)
y=mydata['Close']
#output_notebook()
p = Line(y, title="Closing Price/ Day in May",width=500, height=400,xlabel='Dates', ylabel='Closing Price')
#output_file("templates\closing.html", title="Closing Price in May")
return p
def plot_column(data, column_name, save=True, display=True):
plot = Line(data, y=column_name, plot_width=1200, plot_height=600)
if save:
output_file('plots/' + column_name + '.html')
if display:
show(plot)
return plot
def doPlot11(data, nrDataSource):
p = Line(data,
y_mapper_type="log",
x=data.columns[0],
xlabel=data.columns[0],
ylabel=data.columns[1],
title="Line graph: " + nrDataSource['name'],
responsive=True)
p._xaxis.ticker = SingleIntervalTicker(interval=5, num_minor_ticks=10)
c = components(p, resources=None, wrap_script=False, wrap_plot_info=True)
return c
def trying():
now=datetime.datetime.now().strftime("%Y%m%d")
now_30=(datetime.datetime.now()-datetime.timedelta(days=30)).strftime("%Y%m%d")
tick=app.vars
stock_data = requests.get("https://www.quandl.com/api/v3/datatables/WIKI/PRICES.json?date.gte="+now_30+"&date.lt="+now+"&ticker="+tick+"&api_key=yvSB52TFjUsFTyZU-n--")
unp_data=stock_data.json()["datatable"]["data"]
unp_label=stock_data.json()["datatable"]["columns"]
label=[]
for entries in unp_label:
label.append(str(entries["name"]))
df=pandas.DataFrame(unp_data)
df.columns=label
close_mean="%.2f" %df[["close"]].mean()
close_std="%.2f" %df[["close"]].std()
captions="mean="+ close_mean +", std="+ close_std
df2=df[["date","close"]]
li=Line(df2,x="date",y="close",xlabel="date",ylabel="closing price",legend=False,title=tick+" stock chart from "+now_30+" to "+now)
caption1=Label(x=400, y=100, x_units='screen', y_units='screen', text=captions, text_font_size='9pt', text_font_style="bold")
li.add_layout(caption1)
script, div = components(li)
return render_template('trying.html', script=script, div=div)
def HLevelLine():
# sample = pd.read_table('http://kelesidis.de/static/data/sample.txt')
# Use the Sample Data
# Make a list with the months
ys = [month for month in sample.columns[1:]]
TOOLS = 'pan,wheel_zoom,hover,crosshair,resize,reset'
TOOLTIPS = [("Year", "$~x"),
("Temp", "$y")]
# make our line configurations
# import from df, x= the Year column, y = the ys list we created above
p = Line(sample, x = 'Year', y = ys, title = "Hight Level Bokeh Line Chart", legend = "bottom_left",
ylabel = 'Temp', tools = TOOLS, width = 600, height = 450, responsive = True)
# hover tool configuration
p_hover = p.select(HoverTool)
p_hover.tooltips = TOOLTIPS
p.logo = None
return p
def main():
# check recalculation request
if 'recalculate' in request.args:
if request.args.get('recalculate') == 'True':
betalyzer.recalculate()
# build sector betas bar chart
sector_betas = betalyzer.df_tickers.groupby('sector')['beta'].mean()
bk_sector_betas = Bar(sector_betas,
plot_width=550,
plot_height=400,
legend=None)
bk_sector_betas_script, bk_sector_betas_div = components(bk_sector_betas)
# build market cap betas bar chart
mktcap_betas = betalyzer.df_tickers.groupby(
'market_cap_decile')['beta'].mean()
bk_mc_betas = Bar(mktcap_betas,
plot_width=550,
plot_height=400,
legend=None)
bk_mc_betas_script, bk_mc_betas_div = components(bk_mc_betas)
# build market cap scatter plot
scatter = Scatter(betalyzer.df_tickers,
x='market_cap_log',
y='beta',
plot_width=550,
plot_height=400)
scatter_script, scatter_div = components(scatter)
# build line plot for top three stocks
top_tickers = betalyzer.df_tickers['ticker'].head(3)
bk_history = Line(betalyzer.df_betas[top_tickers],
plot_width=550,
plot_height=400)
bk_history_script, bk_history_div = components(bk_history)
return render_template(
'main.html',
dt_tickers=betalyzer.df_tickers.to_dict(orient='records'),
bk_sector_betas_script=bk_sector_betas_script,
bk_sector_betas_div=bk_sector_betas_div,
bk_mc_betas_script=bk_mc_betas_script,
bk_mc_betas_div=bk_mc_betas_div,
scatter_script=scatter_script,
scatter_div=scatter_div,
bk_history_script=bk_history_script,
bk_history_div=bk_history_div)
def get_variable(config):
from bokeh.embed import components
name, rank, df, decade_df = get_result(config.gender, config.decade, config.name)
if config.name <> name:
message = "%s not found. Do you mean %s?" % (config.name, name)
else:
message =""
birth_table = df.pivot(index='Decade', columns='Name', values='Births').fillna(0).to_html()
rank_table = df.pivot(index='Decade', columns='Name', values='Rank').fillna(0).to_html()
result_table = df[df['Rank']==rank][['Decade','Name','Rank']].sort('Decade').to_html(index=False)
top_table = decade_df[(decade_df['Rank']<=8) &
(decade_df['Decade']==config.decade)].sort('Rank').to_html(index=False)
from bokeh.charts import Line, save, output_file, ColumnDataSource
from bokeh.resources import INLINE
plot_path = "%s_%s_%i.html" % (config.name,config.gender,config.decade)
output_file("output/" + plot_path, mode='inline')
tooltips = [(c, '@' + c) for c in df.columns]
p = Line(df, x='Decade', y='Rank', title="Rank across Time", color='Name',
xlabel="Decade", ylabel="Rank",
tooltips=tooltips)
p.circle('Decade', 'Rank', color='gray', alpha=0.5, source=ColumnDataSource(df))
save(p)
#script, div = components(p)
return {
'plot_path': plot_path,
'result_table': result_table,
'rank_table': rank_table,
'birth_table': birth_table,
'top_table': top_table,
'rank': rank,
'config': config,
'name': name,
'message': message
}
def getPlot(ticker):
ticker = 'GOOG'
r = requests.get('https://www.quandl.com/api/v3/datasets/WIKI/' + ticker +
'/data.json?api_key=Eebg1xGXqxhS11D52xGs')
asJson = json.loads(r.content)
dataFormated = pd.DataFrame(np.array(asJson['dataset_data']['data']),
columns=asJson['dataset_data']['column_names'])
dataFormated['Date'] = pd.to_datetime(dataFormated.Date)
dataFormated['Open'] = pd.to_numeric(dataFormated.Open)
#dates = matplotlib.dates.date2num(asDateTime.tolist())
#values = np.array(dataFormated.Open).astype(float)
line = Line(dataFormated, x='Date', y='Open', title="Cool")
#line = Line(x=dates,y=values,title="Cool")
return line
def plot_plan_bokeh_2(plan):
x = []
y = []
ys = []
i = 0
for (o, v) in plan:
# s = str(o) + " - " + str(v)
s = str(o)
if s not in ys:
ys.append(s)
for (o, v) in plan:
i += 1
x.append(i)
# s = str(o) + " - " + str(v)
s = str(o)
y.append(ys.index(s))
# xyvalues = np.array([[2, 3, 7, 5, 26], [12, 33, 47, 15, 126], [22, 43, 10, 25, 26]])
xyvalues = np.array(y)
line = Line(xyvalues, title="plan", legend="top_left", ylabel='operator')
output_file('plots_html/line.html')
show(line)
def ticker(ticker):
# build line
line = Line(betalyzer.df_betas[ticker], plot_width=1000, plot_height=400)
bokeh_script, bokeh_div = components(line)
# build scatter
scatter = Scatter(betalyzer.df_changes.head(betalyzer.window),
x=betalyzer.market,
y=ticker,
plot_width=550,
plot_height=400)
scatter_script, scatter_div = components(scatter)
# build histogram
df_hist = betalyzer.df_changes[[ticker,
'SPY']].head(500).unstack().reset_index()
df_hist.rename(columns={'level_0': 'ticker', 0: 'change'}, inplace=True)
hist = Histogram(df_hist,
values='change',
color='ticker',
bins=20,
legend='top_right',
plot_width=550,
plot_height=400)
hist_script, hist_div = components(hist)
return render_template(
'ticker.html',
ticker=ticker,
bokeh_script=bokeh_script,
bokeh_div=bokeh_div,
scatter_script=scatter_script,
scatter_div=scatter_div,
hist_script=hist_script,
hist_div=hist_div,
window=betalyzer.window,
dt_ticker=betalyzer.df_tickers.loc[ticker].to_dict())
def generate_graph():
url = 'https://www.quandl.com/api/v3/datatables/WIKI/PRICES.json?'
r = requests.get(url, app.selection)
rjson = r.json()['datatable']
data = pd.DataFrame(rjson['data'])
cols = pd.DataFrame(rjson['columns'])['name']
data.columns = cols
def convert_to_datetime(val):
y, m, d = val.split('-')
return pd.datetime(int(y), int(m), int(d))
data['date'] = data['date'].apply(convert_to_datetime)
data = data.set_index(['date'])
t = 'Time series data for ' + app.selection['ticker']
p = Line(data, title=t, xlabel='date', ylabel='price ($)')
app.s, app.d = components(p)
# html = file_html(p,CDN,'Data sourced from Quandl WIKI dataset')
# f = open('templates/graph.html','w')
# f.write(html)
# f.close()
return
from collections import OrderedDict
import time
import numpy as np
from bokeh.charts import Line, curdoc, cursession, output_server, show
from bokeh.models import GlyphRenderer
N = 80
x = np.linspace(0, 4*np.pi, N)
xyvalues = OrderedDict(sin=np.sin(x), cos=np.cos(x))
output_server("line_animate")
chart = Line(xyvalues, title="Lines", ylabel='measures')
curdoc().add(chart)
show(chart)
renderer = chart.select(dict(type=GlyphRenderer))
ds = renderer[0].data_source
while True:
for i in np.hstack((np.linspace(1, -1, 100), np.linspace(-1, 1, 100))):
for k, values in xyvalues.items():
if k != 'x':
ds.data['y_%s'%k] = values * i
cursession().store_objects(ds)
time.sleep(0.05)
mydata.head()
mydata.plot(figsize=(12, 8))
import seaborn as sns
mydata.plot(figsize=(12, 8))
##BOKEH
from bokeh.plotting import show
from bokeh.io import output_notebook
from bokeh.charts import Line
#doesn't work in spyder!!
output_notebook()
p = Line(mydata, legend='bottom_right')
p.width = 600
p.height = 400
show(p)
#HTML
from bokeh.plotting import figure
from bokeh.palettes import RdYlBu11
from bokeh.plotting import show, output_file, reset_output
output_file("bokehplot.html")
mytools = ['pan', 'box_zoom', 'resize', 'wheel_zoom',
'reset'] # Does not allow save and question options
p = figure(width=800, height=600, tools=mytools, x_axis_type='datetime')
cols = mydata.columns.values[:-1]
def make_plot():
sel_cols = [output_cols[i] for i in chkbx.active]
p = Line(data=df, x='Date', y=sel_cols, plot_height=400)
return p
import pandas as pd
from bokeh.sampledata.degrees import data
defaults.width = 500
defaults.height = 300
TOOLS='box_zoom,box_select,hover,crosshair,resize,reset'
TOOLTIPS = [ ("y", "$~y"), ("x", "$~x") ]
data = data[['Biology', 'Business', 'Computer Science', "Year"]]
data = pd.melt(data, id_vars=['Year'],
value_vars=['Biology', 'Business', 'Computer Science'],
value_name='Count', var_name='Degree')
vline = Line(data, y='Count', color='Degree', title="Lines VLine", ylabel='measures',
tools=TOOLS)
hline = Line(data, y='Count', color='Degree', title="Lines HLine",
ylabel='measures', tools=TOOLS)
int_vline = Line(data, y='Count', color='Degree', title="Lines VLine Interp",
ylabel='measures', tools=TOOLS)
int_hline = Line(data, y='Count', color='Degree', title="Lines HLine Interp",
ylabel='measures', tools=TOOLS)
scatter_point = Scatter(data, x='Year', y='Count', color='Degree',
title="Scatter mouse", ylabel='measures', legend=True,
tools=TOOLS)
scatter = Scatter(data, x='Year', y='Count', color='Degree',
return breath, status, vol
df = pd.read_csv(r'c:\Research_data\DH_data\test-840-1.txt', engine = 'python', skiprows = 2, skipfooter = 2,
names = ['flow', 'paw', 'other'])
df.reset_index(drop = False, inplace = True)
df.drop(labels = 'other', axis = 1, inplace = True)
df.flow = pd.to_numeric(df.flow, errors = 'coerce')
df.paw = pd.to_numeric(df.paw, errors = 'coerce')
df['breath'], df['status'], df['vol'] = count_breath(df.flow.values)
df.dropna(axis = 0, how = 'all', subset = ['flow'], inplace = True)
df.vol = df.groupby('breath')['vol'].cumsum()
p = Line(df, x = 'index', y = ['flow', 'paw', 'vol', 'breath', 'status'], color = 'red')
p.extra_y_ranges = {'flow': Range1d(start = 0, end = 30)}
p.add_layout(LinearAxis(y_range_name = 'flow'), 'left')
check_button = CheckboxButtonGroup(labels = ['Double Stacked', 'Flow Limited', 'Ineffective Trigger'],
active = [0, 0, 0])
next_button = Button(label = 'Next', type = 'success')
next_button.on_click(print('ok'))
output_file('test.html')
show(vform(p, check_button, next_button))
defaults.height = 300
TOOLS = 'box_zoom,box_select,hover,crosshair,resize,reset'
TOOLTIPS = [("y", "$~y"), ("x", "$~x")]
data = data[['Biology', 'Business', 'Computer Science', "Year"]]
data = pd.melt(data,
id_vars=['Year'],
value_vars=['Biology', 'Business', 'Computer Science'],
value_name='Count',
var_name='Degree')
vline = Line(data,
y='Count',
color='Degree',
title="Lines VLine",
ylabel='measures',
tools=TOOLS)
hline = Line(data,
y='Count',
color='Degree',
title="Lines HLine",
ylabel='measures',
tools=TOOLS)
int_vline = Line(data,
y='Count',
color='Degree',
title="Lines VLine Interp",
ylabel='measures',
from bokeh.models import HoverTool
from bokeh.charts import Chart, Step, Line, Area, Scatter, Bar, vplot, hplot, show, output_file
from py import path
HERE = path.local(__file__).dirpath()
xyvalues = pd.read_csv(
str(HERE.join("percent-bachelors-degrees-women-usa.csv")))
index = xyvalues.pop("Year")
xyvalues = xyvalues[['Biology', 'Business', 'Computer Science']]
TOOLS = 'box_zoom,box_select,hover,crosshair,resize,reset'
output_file("lines.html", title="line.py example")
vline = Line(xyvalues,
title="Lines VLine",
ylabel='measures',
width=500,
height=300,
tools=TOOLS)
hline = Line(xyvalues,
title="Lines HLine",
ylabel='measures',
width=500,
height=300,
tools=TOOLS)
int_vline = Line(xyvalues,
title="Lines VLine Interp",
ylabel='measures',
width=500,
height=300,
tools=TOOLS)
int_hline = Line(xyvalues,
from collections import OrderedDict
import numpy as np
import pandas as pd
from bokeh.charts import Line
xyvalues = OrderedDict(
python=[2, 3, 7, 5, 26, 221, 44, 233, 254, 265, 266, 267, 120, 111],
pypy=[12, 33, 47, 15, 126, 121, 144, 233, 254, 225, 226, 267, 110, 130],
jython=[22, 43, 10, 25, 26, 101, 114, 203, 194, 215, 201, 227, 139, 160],
)
# any of the following commented are valid Line inputs
#xyvalues = pd.DataFrame(xyvalues)
#xyvalues = xyvalues.values()
#xyvalues = np.array(xyvalues.values())
line = Line(xyvalues, title="Lines", ylabel='measures', filename="lines.html")
line.xlabel('time').legend("top_left").show()
import numpy as np
import pandas as pd
from bokeh.models import HoverTool
from bokeh.charts import Chart, Step, Line, Area, Scatter, Bar, vplot, hplot, show, output_file
from py import path
HERE = path.local(__file__).dirpath()
xyvalues = pd.read_csv(str(HERE.join("percent-bachelors-degrees-women-usa.csv")))
index = xyvalues.pop("Year")
xyvalues = xyvalues[['Biology', 'Business', 'Computer Science']]
TOOLS='box_zoom,box_select,hover,crosshair,resize,reset'
output_file("lines.html", title="line.py example")
vline = Line(xyvalues, title="Lines VLine", ylabel='measures', width=500, height=300,
tools=TOOLS)
hline = Line(xyvalues, title="Lines HLine", ylabel='measures', width=500, height=300,
tools=TOOLS)
int_vline = Line(xyvalues, title="Lines VLine Interp", ylabel='measures', width=500, height=300,
tools=TOOLS)
int_hline = Line(xyvalues, title="Lines HLine Interp", ylabel='measures', width=500, height=300,
tools=TOOLS)
svalues = {}
# svalues['Business'] = [(i, v) for i, v in zip(index, xyvalues['Business'])]
for k in xyvalues.columns:
svalues[k] = [(i, v) for i, v in zip(index, xyvalues[k])]
# # import pdb; pdb.set_trace()
scatter_point = Scatter(svalues, title="Scatter mouse", ylabel='measures', width=500, height=300,
legend=True,
tools=TOOLS)
scatter = Scatter(svalues, title="Scatter V Line", ylabel='measures', width=500, height=300,
# build a dataset where multiple columns measure the same thing
data = dict(
python=[2, 3, 7, 5, 26, 221, 44, 233, 254, 265, 266, 267, 120, 111],
pypy=[12, 33, 47, 15, 126, 121, 144, 233, 254, 225, 226, 267, 110, 130],
jython=[22, 43, 10, 25, 26, 101, 114, 203, 194, 215, 201, 227, 139, 160],
test=[
'foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'bar',
'foo', 'bar', 'foo', 'bar'
])
df = pd.DataFrame(data)
# add a column with a range of dates, as if the values were sampled then
df['date'] = pd.date_range('1/1/2015', periods=len(df.index), freq='D')
# default behavior for dataframe input is to plot each numerical column as a line
line = Line(df)
# build the line plots
line0 = Line(df,
y=['python', 'pypy', 'jython'],
title="Interpreters (y=['python', 'pypy', 'jython'])",
ylabel='Duration',
legend=True)
line1 = Line(df,
x='date',
y=['python', 'pypy', 'jython'],
title="Interpreters (x='date', y=['python', 'pypy', 'jython'])",
ylabel='Duration',
legend=True)
def do_single_search(request_form):
"""
search method called from both welcome() and search()
:param request_form:
:return:
"""
search_terms = request_form["singleTermQuery"].lower()
language_var, country_var = request_form["languageAndRegion"].split(':', 1)
try:
specific_query = simple_query_totals({"query": "body_text_ws:%s" % search_terms,
"filter": ["country_s:%s" % country_var, "langid_s:%s" % language_var]})
except (KeyError, HTTPError):
return flask.render_template('no_results.html', query=search_terms, available_options=AVAILABLE_OPTIONS,
search_mode='single')
matches = specific_query['num_docs'].sum()
#############################
# GET TOTALS FOR EVERYTHING #
#############################
totals = simple_query_totals({"query": "*:*",
"filter": ["country_s:%s" % country_var, "langid_s:%s" % language_var]})
gender_totals = totals.groupby('gender').num_docs.sum()
age_totals = totals.groupby('age').num_docs.sum()
age_totals = sort_and_filter_age(age_totals)
age_totals_norm = age_totals / age_totals.sum()
age_and_gender_totals = prepare_age_and_gender(totals)
# nuts_total = totals.groupby('nuts_3').num_docs.sum()
###########
# GENDER #
###########
gender_specific_query = specific_query.groupby('gender').num_docs.sum()
abs_percentages = gender_specific_query / gender_totals
try:
renormalizer = 1.0 / abs_percentages.sum()
except ZeroDivisionError:
return flask.render_template('no_results.html', query=search_terms, available_options=AVAILABLE_OPTIONS,
search_mode='single')
gender_query_adjusted = abs_percentages * renormalizer
#######
# AGE #
#######
age_specific_query = specific_query.groupby('age').num_docs.sum()
age_specific_query = sort_and_filter_age(age_specific_query)
age_specific_query_norm = age_specific_query / age_specific_query.sum()
compare_age_df = pd.DataFrame({'background distribution': age_totals_norm,
'query': pd.rolling_mean(age_specific_query_norm, ROLLING_MEAN_FRAME)})
compare_age_df['i'] = compare_age_df.index
##################
# AGE AND GENDER #
##################
age_and_gender_specific_query = prepare_age_and_gender(specific_query)
try:
age_specific_male_totals = gender_specific_query['M'].sum()
compare_male_df = pd.DataFrame({'background distribution': age_and_gender_totals['M'],
'query': pd.rolling_mean(age_and_gender_specific_query['M'],
ROLLING_MEAN_FRAME)})
except KeyError:
age_specific_male_totals = 0
compare_male_df = pd.DataFrame({'background distribution': age_and_gender_totals['M']})
compare_male_df['i'] = compare_male_df.index
try:
age_specific_female_totals = gender_specific_query['F']
compare_female_df = pd.DataFrame({'background distribution': age_and_gender_totals['F'],
'query': pd.rolling_mean(age_and_gender_specific_query['F'],
ROLLING_MEAN_FRAME)})
except KeyError:
age_specific_female_totals = 0
compare_female_df = pd.DataFrame({'background distribution': age_and_gender_totals['F']})
compare_female_df['i'] = compare_female_df.index
########
# NUTS #
########
nuts_query = specific_query.groupby('nuts_3').num_docs.sum()
nuts_total = nuts_query.sum()
nuts_query_norm = nuts_query / nuts_total
special_regions = nuts_query_norm > nuts_query_norm.median()
outliers = ', '.join(
sorted(['%s (%s)' % (NUTS_NAMES[x], x) for x in special_regions.index if special_regions.ix[x].any() == True]))
# TODO move plotting to its own function
gender_plot = Bar(gender_query_adjusted,
title="Gender distribution",
ylabel="percentage",
logo=None,
toolbar_location="below",
# width=300,
# height=400,
webgl=False)
age_plot = Line(compare_age_df,
x='i',
title="Age distribution",
x_range=Range1d(start=MIN_AGE, end=MAX_AGE),
xlabel='age',
ylabel="percentage",
logo=None,
toolbar_location="below",
# width=800,
# height=400,
legend='top_right',
color=['silver', 'red'],
webgl=False)
age_gender_plot_M = Line(compare_male_df,
x='i',
title="Age distribution for men",
xlabel='age',
ylabel="percentage",
x_range=Range1d(start=MIN_AGE, end=MAX_AGE),
logo=None,
toolbar_location="below",
# width=600,
# height=400,
legend='top_right',
color=['silver', 'green'],
webgl=False)
age_gender_plot_F = Line(compare_female_df,
x='i',
title="Age distribution for women",
xlabel='age',
x_range=Range1d(start=MIN_AGE, end=MAX_AGE),
logo=None,
toolbar_location="below",
# width=600,
# height=400,
legend='top_right',
color=['silver', 'blue'],
webgl=False)
bokeh_script, (gender_plot_div, age_plot_div, age_gender_plot_F_div, age_gender_plot_M_div) = components(
(gender_plot, age_plot, age_gender_plot_F, age_gender_plot_M))
return flask.render_template('single_term_results.html',
query=search_terms,
matches=matches,
bokeh_script=bokeh_script,
gender_query_adjusted=gender_query_adjusted,
gender_plot=gender_plot_div,
age_plot=age_plot_div,
age_gender_plot_F=age_gender_plot_F_div,
age_gender_plot_M=age_gender_plot_M_div,
country_code=country_var,
map_views=MAP_VIEWS,
nuts_query=nuts_query_norm.to_json(),
outliers=outliers,
gender_total=gender_specific_query.sum(),
age_total=age_specific_query.sum(),
age_total_M=age_specific_male_totals,
age_total_F=age_specific_female_totals,
nuts_total=nuts_query.sum(),
available_options=AVAILABLE_OPTIONS)
from bokeh.charts import Line, show, output_file
# build a dataset where multiple columns measure the same thing
data = dict(
python=[2, 3, 7, 5, 26, 221, 44, 233, 254, 265, 266, 267, 120, 111],
pypy=[12, 33, 47, 15, 126, 121, 144, 233, 254, 225, 226, 267, 110, 130],
jython=[22, 43, 10, 25, 26, 101, 114, 203, 194, 215, 201, 227, 139, 160],
test=[
'foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'bar',
'foo', 'bar', 'foo', 'bar'
])
# create a line chart where each column of measures receives a unique color and dash style
line = Line(data,
y=['python', 'pypy', 'jython'],
dash=['python', 'pypy', 'jython'],
color=['python', 'pypy', 'jython'],
title="Interpreter Sample Data",
ylabel='Duration',
legend=True)
output_file("line_single.html", title="line_single.py example")
show(line)
# add a column with a range of dates, as if the values were sampled then
df['date'] = pd.date_range('1/1/2015', periods=len(df.index), freq='D')
# default behavior for dataframe input is to plot each numerical column as a line
line = Line(df)
# build the line plots
line0 = Line(df, y=['python', 'pypy', 'jython'],
title="Interpreters (y=['python', 'pypy', 'jython'])", ylabel='Duration', legend=True)
line1 = Line(df, x='date', y=['python', 'pypy', 'jython'],
title="Interpreters (x='date', y=['python', 'pypy', 'jython'])", ylabel='Duration', legend=True)
line2 = Line(df, x='date', y=['python', 'pypy', 'jython'],
dash=['python', 'pypy', 'jython'],
title="Interpreters (x='date', y, dash=['python', 'pypy', 'jython'])", ylabel='Duration', legend=True)
line2.title_text_font_size = '11pt'
line3 = Line(df, x='date', y=['python', 'pypy', 'jython'],
dash=['python', 'pypy', 'jython'],
color=['python', 'pypy', 'jython'],
title="Interpreters (x='date', y, dash, color=['python', 'pypy', 'jython'])", ylabel='Duration', legend=True)
line3.title_text_font_size = '11pt'
line4 = Line(df, x='date', y=['python', 'pypy', 'jython'],
dash='test',
color=['python', 'pypy', 'jython'],
title="Interpreters (x='date', y, color=['python', 'pypy', 'jython'], dash='test') with tooltips", ylabel='Duration',
legend=True, tooltips=[('series', '@series'), ('test', '@test')])
