def layout():
date_select = DateRangeSlider(name="period",
title="Period:",
value=(start, end),
bounds=(bounds_start, bounds_end),
value_labels='show',
range=(dict(days=1), None))
date_select.on_change('value', on_date_change)
country_select = Select(title="Host Site Country:",
value="World",
options=country_choices)
country_select.on_change('value', on_country_change)
controls = VBoxModelForm(_children=[
Paragraph(text="Date Range"),
Paragraph(text=""), # spacing hack
Paragraph(text=""),
date_select,
country_select
])
vboxsmall = VBoxModelForm(_children=[controls, source_par])
#hbox1 = HBox(children=[job_loc_plot_builder(), vboxsmall])
#hbox2 = HBox(children=[weekday_builder(), jobtype_builder()])
#layout = VBox(children=[hbox1, hbox2])
layout = VBox(children=[vboxsmall])
return layout
def assemble_precision_recall_tab(self):
dropdown_value = self.model_dropdown.value
threshold_value = self.threshold_slider.value
precision_and_recall = self.daily_precision_and_recall_for_threshold(
threshold_value, self.model_probas[dropdown_value],
self.test_labels)
daily_metrics = self.daily_metrics[dropdown_value]
vis_x_axis = daily_metrics.index
min_date, max_date = vis_x_axis.min(), vis_x_axis.max()
time_period_slider = DateRangeSlider(title="Time Period",
value=(min_date, max_date),
start=min_date,
end=max_date,
step=1)
precision_recall_f = figure(plot_width=700,
plot_height=400,
x_axis_type="datetime",
title=dropdown_value)
precision_recall_f.xaxis.axis_label = "Date"
precision_recall_f.yaxis.axis_label = "Precision"
prec_source = ColumnDataSource(precision_and_recall["precision"])
rec_source = ColumnDataSource(precision_and_recall["recall"])
precision_recall_f.line("date",
"precision",
source=prec_source,
color="orange",
legend="Precision")
precision_recall_f.line("date",
"recall",
source=rec_source,
color="green",
legend="Recall")
precision_recall_f.legend.location = "bottom_right"
def update(attr, old, new):
model = self.model_dropdown.value
threshold = self.threshold_slider.value
date_start, date_end = time_period_slider.value_as_datetime
precision_recall_f.title.text = model
precision_and_recall = self.daily_precision_and_recall_for_threshold(
threshold, self.model_probas[model], self.test_labels)
prec_metrics = precision_and_recall["precision"][
date_start:date_end]
recall_metrics = precision_and_recall["recall"][
date_start:date_end]
prec_source.data = ColumnDataSource.from_df(prec_metrics)
rec_source.data = ColumnDataSource.from_df(recall_metrics)
self.model_dropdown.on_change('value', update)
time_period_slider.on_change('value', update)
self.threshold_slider.on_change('value', update)
layout = column(time_period_slider, precision_recall_f)
return layout
def Create_Range_Sliders():
range_slider_x = RangeSlider(title='X-Axis Range',
start=0,
end=1,
value=(0, 1),
step=0.1)
range_slider_y = RangeSlider(title='Y-Axis Range',
start=0,
end=1,
value=(0, 1),
step=0.1)
range_slider_xdate = DateRangeSlider(title='X-Axis Range (Date)',
start=date(2017, 1, 1),
end=date(2017, 1, 2),
value=(date(2017, 1,
1), date(2017, 1, 2)),
step=1)
range_slider_ydate = DateRangeSlider(title='Y-Axis Range (Date)',
start=date(2017, 1, 1),
end=date(2017, 1, 2),
value=(date(2017, 1,
1), date(2017, 1, 2)),
step=1)
return (range_slider_x, range_slider_y, range_slider_xdate,
range_slider_ydate)
def layout():
date_select = DateRangeSlider(
name="period",
title="Period:",
value=(start, end),
bounds=(bounds_start, bounds_end),
value_labels='show',
range=(dict(days=1), None))
date_select.on_change('value', on_date_change)
country_select = Select(title="Host Site Country:", value="World",
options=country_choices)
country_select.on_change('value', on_country_change)
controls = VBoxModelForm(_children=[Paragraph(text="Date Range"),
Paragraph(text=""), # spacing hack
Paragraph(text=""),
date_select, country_select])
vboxsmall = VBoxModelForm(_children=[controls, source_par])
#hbox1 = HBox(children=[job_loc_plot_builder(), vboxsmall])
#hbox2 = HBox(children=[weekday_builder(), jobtype_builder()])
#layout = VBox(children=[hbox1, hbox2])
layout = VBox(children=[vboxsmall])
return layout
def plot(df, feauture):
def create_plot(source,country, feauture):
hover = HoverTool(tooltips=[("Date", "@date{%F}"), (feauture, str('@'+feauture))],
formatters = {'@date':'datetime'})
p = figure(plot_width=900, plot_height=500,x_axis_type="datetime",
title="Total Covid-19 {} in {} by date".format(feauture, country),
toolbar_location='above')
p.line(x='date', y=feauture, line_width=1, source=source,
line_color="black")
p.vbar(x='date', top=feauture, line_width=1, source=source, fill_color='orange', hover_fill_color='grey',
hover_fill_alpha=1)
p.add_tools(hover)
p.xaxis.axis_label = 'Date'
p.yaxis.axis_label = 'Number of {}'.format(feauture)
p.axis.axis_label_text_font_style = 'bold'
p.yaxis.formatter = NumeralTickFormatter(format = '0.0a')
p.background_fill_color = "beige"
return p
def update(attr, old, new):
country = country_select.value
new_df = df[df['location'] == country]
start = datetime.fromtimestamp(date_select.value[0] / 1000)
end = datetime.fromtimestamp(date_select.value[1] / 1000)
new_df = new_df[new_df['date'] >= start]
new_df = new_df[new_df['date'] <= end]
new_src = ColumnDataSource(new_df)
source.data.update(new_src.data)
p.title.text = 'Total covid-19 {} in {}'.format(feauture, country)
countries = list(df['location'].unique())
country_select = Select(title = 'Country', value = 'world', options = countries)
country_select.on_change('value', update)
date_select = DateRangeSlider(title = 'Date Range', start = min(df['date']), end = max(df['date']),
value = (min(df['date']), max(df['date'])), step = 15)
date_select.on_change('value', update)
initial_country = country_select.value
source = ColumnDataSource(df[df['location'] == initial_country])
p = create_plot(source, initial_country, feauture)
controls = Column(country_select, date_select)
layout = row(controls, p)
return layout
def build_widgets_wb(stock_list):
# CODE SECTION: setup widgets, widgetbox name = controls_wb
WIDGET_WIDTH = 250
# ========== Select Stocks ============= #
select_stk_1 = Select(width = WIDGET_WIDTH, title='Select Stock 1:', value = stock_list[0], options=stock_list)
select_stk_2 = Select(width = WIDGET_WIDTH, title='Select Stock 2:', value = stock_list[0], options=stock_list)
# ========== Strategy Type ============= #
strategy_list = ['kalman', 'distance', 'cointegration']
select_strategy = Select(width = WIDGET_WIDTH, title='Select Strategy:', value = strategy_list[0], options=strategy_list)
# ========== set start/end date ============= #
# date time variables
MAX_START = date(2014, 1, 1)
MAX_END = date(2018, 12, 30)
DEFAULT_START = date(2016, 5, 1)
DEFAULT_END = date(2016, 12, 31)
STEP = 1
backtest_dates = DateRangeSlider(width = WIDGET_WIDTH,
start=MAX_START, end=MAX_END,
value=(DEFAULT_START, DEFAULT_END),
step=STEP, title="Backtest Date Range:")
start_bt = Button(label="Backtest", button_type="success", width = WIDGET_WIDTH)
# controls = column(select_stk_1, select_stk_2, select_strategy, backtest_dates, start_bt)
controls_wb = widgetbox(select_stk_1, select_stk_2, select_strategy, backtest_dates, start_bt, width=600)
return controls_wb
def date_slider(p, *, dates=None, date_column='dt'):
# todo a bit crap, but kinda works.. not sure what's the proper way?
# maybe they aren't computed before show()??
# p.x_range.on_change('start', lambda attr, old, new: print("Start", new))
# p.x_range.on_change('end', lambda attr, old, new: print("End", new))
if dates is None:
graphs = p.renderers
# todo autodetect by type instead of name?
dts = np.concatenate([g.data_source.data[date_column] for g in graphs])
# FFS... apparently no easier way
sdate = pd.Timestamp(np.min(dts)).to_pydatetime()
edate = pd.Timestamp(np.max(dts)).to_pydatetime()
else:
sdate = min(dates)
edate = max(dates)
# todo not sure if should use today?
edate += timedelta(days=5)
from bokeh.models.widgets import DateRangeSlider # type: ignore
ds = DateRangeSlider(
title="Date Range: ",
start=sdate,
end=edate,
value=(sdate, edate),
step=1,
)
from bokeh.models import CustomJS # type: ignore
# TODO hmm. so, js won't be able to call into python in Jupyter...
# see https://docs.bokeh.org/en/latest/docs/gallery/slider.html
update_js = CustomJS(args=dict(ds=ds, xrange=p.x_range),
code=J('''
const [ll, rr] = ds.value;
// didn't work??
// xrange.set({"start": ll, "end": rr})
xrange.start = ll;
xrange.end = rr;
// todo hmm, turned out it wasn't necessary??
// source.trigger('change');
'''))
# todo add some quick selectors, e.g. last month, last year, etc like plotly
ds.js_on_change('value', update_js)
return ds
def source_filters(sources):
def _handle_state_filter():
for fullname in sources['dataframes']:
df = sources['dataframes'][fullname]
df = df[(df.index >= _state['date_range'][0]) & (df.index <= _state['date_range'][1])]
units = UNIT_CONVERSION_MAP[_state['units']]
df = df * UNIT_CONVERSION[units]
sources['sources'][fullname].data = ColumnDataSource(df).data
for callback in ELEMENT_CALLBACKS:
callback(sources=sources, **_state)
def _date_slider_handler(attr, old, new):
start_time, end_time = new
start_date = dt.datetime.fromtimestamp(start_time / 1000)
end_date = dt.datetime.fromtimestamp(end_time / 1000)
_state['date_range'] = (start_date, end_date)
_handle_state_filter()
start_date = dt.datetime(year=2018, month=12, day=1)
end_date = dt.datetime(year=2018, month=12, day=1)
date_slider = DateRangeSlider(start=_state['date_range'][0], end=_state['date_range'][1], value=_state['date_range'], step=1, title="Date Range", height=100)
date_slider.on_change('value', _date_slider_handler)
def _radio_button_group_handler(attr, old, new):
_state['units'] = int(new)
_handle_state_filter()
radio_button_group = RadioButtonGroup(labels=UNIT_CONVERSION_MAP, active=0, width=700, height=100)
radio_button_group.on_change('active', _radio_button_group_handler)
div_button_group_selection = Div(text=f'<img src="/visualization/static/images/{UNIT_CONVERSION_MAP[_state["units"]]}.svg" height="100px"/>', height=100)
def _update_div_button_group(**kwargs):
units = UNIT_CONVERSION_MAP[kwargs['units']]
div_button_group_selection.text = f'<img src="/visualization/static/images/{units}.svg" height="100px"/>'
ELEMENT_CALLBACKS.append(_update_div_button_group)
return [date_slider, radio_button_group, div_button_group_selection]
def sentiment_viz():
#Imports
import mysql.connector
import sqlalchemy as db
import pandas as pd
from bokeh.plotting import figure, output_file, show
from bokeh.models.widgets import DateRangeSlider,TextInput,Dropdown
from bokeh.models.widgets.inputs import AutocompleteInput
from bokeh.models import Panel, Tabs
from bokeh.layouts import column
from bokeh.embed import components
import numpy as np
import datetime
#query mysql for article data and aggregate by day
con = db.create_engine('mysql+mysqlconnector://*******************')
df = pd.read_sql('SELECT * FROM articles', con=con)
df['publishedAt'] = pd.to_datetime(df['publishedAt'])
df['date_only']=df['publishedAt'].apply(lambda x: x.date())
dailycounts=df.groupby(by='date_only').agg({'preds':['count','mean']})
dailycounts.columns=['volume','sentiment_net']
dailycounts['date'] = dailycounts.index
publist = df['source_name'].unique().tolist()
pollist = [] #pd.read_sql('SELECT * FROM politicians',con=con)
#Sentiment Valence Chart
pnet = figure(plot_width=1000, plot_height=300, x_axis_type='datetime')
pnet.toolbar.logo = None
pnet.toolbar_location = None
pnet.background_fill_color = "lavender"
#News Volume Chart
pvol = figure(plot_width=1000, plot_height=100, x_axis_type='datetime')
pvol.toolbar.logo = None
pvol.toolbar_location = None
pvol.background_fill_color = "bisque"
#Render data lines
pnet.line(dailycounts['date'],
dailycounts['sentiment_net'],
line_width=2,
color='purple')
pvol.line(dailycounts['date'],
dailycounts['volume'],
line_width=2,color='orange')
#Menu objects
polmenu = AutocompleteInput(completions=pollist,title='Politician (case sensitive)')
pubmenu = AutocompleteInput(completions=publist,title='Publication (case sensitive)')
date_slider = DateRangeSlider(start=min(dailycounts['date']), end=max(dailycounts['date']),
value=(min(dailycounts['date']),max(dailycounts['date'])), step=1, title="Date Range")
#create components to be rendered in html
script,div = components(column(pnet,pvol,polmenu,pubmenu,date_slider))
print(dailycounts.head())
return script,div
def month_range_options(self):
"""Returns gridplot (bokeh layout or Element) defining elements for the User to manipulate Data shown in
other views.
Function first checks if the months variables are initialized and if not, initializes them. Then one
bokeh widget - DateRangeSlider is created.
DateRangeSlider will allow User to filter data to Months between two "borders" of the Slider. Even though
DateRangeSlider defines step: 1 (so all days from the months are shown), formatter applied will only show
month-year values ("%b-%Y).
Additionally, defined callback checks months of start-stop values and compares them to months from
old values. This way, callbacks aren't triggered for every slight change in the Slider and change in
.chosen_months is only triggered when the actual change in months happens.
Values for the Slider are also taken from instance attributes - this way User's previous choice is
remembered and can be shown upon coming back to Settings View.
Returns DateRangeSlider.
"""
if self.is_month_range_initialized is False:
self.__initialize_months()
sld = DateRangeSlider(start=self.all_months[0], end=self.all_months[-1], step=1,
value=(self.chosen_months[0], self.chosen_months[-1]),
format="%b-%Y", title="Chosen Month Range: ",
css_classes=["month_range_slider"])
def month_range_callback(attr, old, new):
formatting = "%Y-%m"
old_str = self.__create_timetuple_string_from_timestamp(old, formatting)
new_str = self.__create_timetuple_string_from_timestamp(new, formatting)
if old_str != new_str:
self.__update_chosen_months(new)
sld.on_change("value", month_range_callback)
return sld
def bokeh_main_2():
import sys
src = Path(__file__).absolute().parent / 'src'
sys.path.insert(0, str(src))
from bokeh.layouts import layout # type: ignore
from bokeh.models.widgets import Tabs, Panel # type: ignore
from bokeh.plotting import figure # type: ignore
from dashboard.core.bokeh import test_scatter_matrix
import holoviews as hv
f1 = hv.render(test_scatter_matrix())
f2 = figure()
f2.circle([0, 0, 2], [4, -1, 1])
# todo need something more lazy?
from dashboard.data import sleep_dataframe
from dashboard.sleep import plot_sleep
# TODO would be cool to display logs in the frontend and also currently evaluated function?
# although pehaps it's easier to just always keep logs open?
from bokeh.models.widgets import DateRangeSlider # type: ignore
from datetime import date
drs = DateRangeSlider(
title="Date Range: ",
start=date(2017, 1, 1),
end=date.today(),
value=(date(2017, 9, 7), date(2017, 10, 15)),
step=1,
)
def update(attr, old, new):
print(attr, old, new)
from bokeh.models import CustomJS # type: ignore
# todo see https://docs.bokeh.org/en/latest/docs/gallery/slider.html
update_js = CustomJS(args=dict(drs=drs),
code='''
console.log("HIIII");
''')
drs.on_change('value', update)
drs.js_on_change('value', update_js)
l1 = layout([[f1, drs]], sizing_mode='fixed')
l2 = layout([[f2]], sizing_mode='fixed')
tabs = Tabs(tabs=[
Panel(child=l1, title='This is Tab 1'),
Panel(child=l2, title='This is Tab 2'),
# Panel(child=layout([[sp]]), title='Sleep dataframe'),
])
curdoc().add_root(tabs)
def assemble_global_widgets(self):
model_menu = [(self.rf_clf.__name__, self.rf_clf.__name__), None,
(self.xgb_clf.__name__, self.xgb_clf.__name__)]
model_dropdown = Dropdown(label="Model",
button_type="warning",
menu=model_menu,
value=self.xgb_clf.__name__)
vis_x_axis = self.test
min_date, max_date = vis_x_axis.index.min(), vis_x_axis.index.max()
time_period_slider = DateRangeSlider(title="Time Period",
value=(min_date, max_date),
start=min_date,
end=max_date,
step=1)
return model_dropdown, time_period_slider
p.title.text = 'Covid-19 Cases and Deaths for ' + sel_country.value + ': ' + sel_chart.value
src.data.update(new_src.data)
p.extra_y_ranges['Avg'].start = 0.95 * np.min(
[src.data['avgc'], src.data['avgd']])
p.extra_y_ranges['Avg'].end = 1.05 * np.max(
[src.data['avgc'], src.data['avgd']])
# set up the controls
sel_country = Select(value="India", options=sorted_by_cases, width=220)
sel_chart = Select(value='Day by Day',
options=['Day by Day', 'Cumulative'],
width=120)
dateslider = DateRangeSlider(start=datemin,
end=datemax,
value=(datemin, datemax),
title='Date Range',
sizing_mode="scale_width")
# set up the event handlers
sel_chart.on_change('value', update)
sel_country.on_change('value', update)
dateslider.on_change('value', update)
# create the dataset and the plot
src = make_dataset(sel_country.value,
sel_chart.value,
range_start=datemin,
range_end=datemax)
p = make_plot(src)
for (i=0; i<source.data.date.length; i++){
if (!source.data.offset_applied[i]){
source.data.date[i] = source.data.date[i] - utcOffset;
}
}
console.log(source.data);
""")
data_src.js_on_change('stream', source_js_callback)
btn = Button(label='Update')
btn.on_click(update_data_src)
date_widget = DateRangeSlider(
title='Date Range',
value=DT_RANGE,
start=DT_RANGE[0],
end=DT_RANGE[1],
)
def source_py_callback(attr, old, new):
if not len(data_src.data['date']):
return
min_dt = min(data_src.data['date'])
max_dt = max(data_src.data['date'])
date_widget.start = min_dt
date_widget.end = max_dt
date_widget.value = (min_dt, max_dt)
data_src.on_change('data', source_py_callback)
words_not_in_vocab = PreText(text='Query words not in vocab: ' +
', '.join(key_list_nonvocab_words))
relevance_threshold = Slider(title="relevance threshold",
value=0.5,
start=0,
end=1,
step=0.01)
app_select = Select(title='Select app:', value=app_names[0], options=app_names)
limit_to_keyword = TextInput(title="limit to reviews containing the word:",
value='')
reviews = PreText(text='', width=1000)
#date_range = RangeSlider(title="Select range of dates", start = 1, end = len(unique_dates), step=1, value=(1, len(unique_dates)), width=1000)
date_range = DateRangeSlider(title="Select range of dates",
start=unique_dates[0],
end=unique_dates[-1],
step=1,
value=(unique_dates[0], unique_dates[-1]),
width=1000)
update_printed_reviews_button = Button(
label="Get relevant reviews in selected timeframe", button_type="success")
def update_app(attrname, old, new):
app_name = app_select.value
unique_dates = unique_dates_dict[app_name]
#just to reset the slider
date_range.start = unique_dates[0]
date_range.end = unique_dates[-1]
date_range.value = (unique_dates[0], unique_dates[-1])
value=0,
step=10)
cp_btnGroup = RadioButtonGroup(name="cp_btngroup",
labels=["less than", "disable", "greater than"],
active=1,
margin=(0, 0, 20, 0))
cp_slider = Slider(
title="Closing Price Change Ratio (%)",
start=-100,
end=100,
value=0,
step=10,
)
date_range_slider = DateRangeSlider(title="Date range",
value=(date_form(data_date[20]),
date_form(data_date[-20])),
start=date_form(data_date[0]),
end=date_form(data_date[-1]),
step=1)
stock_code_input = TextInput(title="Stock Code")
stock_name_input = TextInput(title="Stock Name")
stock_selector = Select(title="Choose Stock to Plot")
mb_source = ColumnDataSource(data={"date": [], "price": []})
cp_source = ColumnDataSource(data={"date": [], "price": []})
TOOLTIPS = [("Title", "@title"), ("Date", "@date_str"), ("Price", "@price")]
p = figure(x_axis_type="datetime",
plot_height=500,
plot_width=600,
title="",
toolbar_location=None,
def first_tab_create(filterData):
########method1: create data source for plots
# dummy data that will be replaced by button values once we get those implemented (right now only granulaity button is implemented)
all_min_date = filterData.groupby('dataid').agg(min)["time"]
all_max_date = filterData.groupby('dataid').agg(max)["time"]
dummy_daterange = ['2019-05-01', '2019-08-20']
dummy_home_id = 27
dummy_data_type = 'car1'
dummy_granularity = '15 Minutes'
def plot1_data(house, daterange=dummy_daterange, data=dummy_data_type, xaxis=dummy_granularity):
# house is an integer number ex. 27
# daterange is an array with 2 strings, start date and end date. ex. ['2019-05-01','2019-08-09']
# weekdays is an ordered list 0-6 of integers ex. [1,4,6] (these are the days we want to exclude)
# data is a string ex. 'car1'
# xaxis is also a string ex. 'hour'
houseData = filterData[filterData['dataid'] == house].sort_values('time', ascending=True)[[data, 'time']]
# that cuts the house, sorts by ascending time, and pulls out only the type of data that was requested
houseData.index = houseData['time'] # reindex by the datetime
houseData = houseData.loc[daterange[0]:daterange[1], :] # cut to the days requested
# Now we get into the xaxis user options #
if xaxis == '15 Minutes':
houseData = houseData.drop(columns="time")
if xaxis == 'Hour':
houseData = houseData.resample('1h').mean()
houseData[data]=houseData[data]*3600/3600
if xaxis == 'Day':
houseData = houseData.resample('1d').mean()
houseData[data] = houseData[data] * 3600 / 3600 * 24
if xaxis == 'Week':
houseData = houseData.resample('1w').mean()
houseData[data] = houseData[data] * 3600 / 3600 *24 * 7
if xaxis == 'Month':
houseData = houseData.resample('1m').mean()
houseData[data] = houseData[data] * 3600 / 3600 *24 * 7 * 30 ############this computation is wrong because n stadard month but just go with it for now
# if none of these, 15 Minutes is implied and passed through
return ColumnDataSource(houseData)
def plot2_data(house,daterange=dummy_daterange,weekdays = [],data=dummy_data_type,xaxis=dummy_granularity):
houseData = filterData[filterData['dataid'] == house].sort_values('time', ascending = True)[[data,'time']]
# that cuts the house, sorts by ascending time, and pulls out only the type of data that was requested
houseData.index = houseData['time'] # reindex by the datetime
houseData = houseData.loc[daterange[0]:daterange[1],:] # cut to the days requested
for i in weekdays:
houseData = houseData[houseData['time'].dt.dayofweek != i] # cut out days we dont want
if xaxis == 'avgday':
houseData = houseData.resample('1d').sum() ####chjange to mean
houseData['time'] = houseData.index
houseData = houseData.groupby(houseData['time'].dt.dayofweek)[data].mean()
houseData.index = ['Monday','Tuesday','Wednesday','Thursday','Friday','Saturday','Sunday']
if xaxis == 'avghour':
houseData = houseData.resample('1h').mean()
houseData['time'] = houseData.index
houseData = houseData.groupby(houseData['time'].dt.hour)[data].mean() # Does not account for UTC change!
houseData = pd.DataFrame(data = houseData)
houseData['axis'] = houseData.index
return ColumnDataSource(houseData)
# now we need to set up our plot axis
##########method2: create plots
def plot1_plot(src):
# create the plot every time a change occurs
plot1 = figure(title="Energy Consumption Per Period", x_axis_type="datetime", x_axis_label="Date",
y_axis_label="Energy Consumption [kWh]")
plot1.line('time', "grid", source=src, ) # simple line plot #data_type_available[data_type_selector.active]
return plot1 # plot object type
def plot2_plot(src):
plot2 = figure(x_range = (0,23), y_range=(-7,2), title='Average Power Consumption For Hours in a Day') # gotta do the ranges as mins and maxes
plot2.vbar(x='axis', top = 'grid', width=1, source=src)
plot2.yaxis.axis_label = 'Power Consumption [kW]'
plot2.xaxis.axis_label = 'Hour of the Day'
#plot2 = figure(title = 'PLOT2')
#plot2.line('axis','grid', source = src)
return plot2
#########Method3: Update App
def update(attr, old, new): # still a little unsure how the update function gets values passed in implicitly
# these values to be replaced with button/user inputs
home_id_to_plot = 27
daterange_to_plot = ['2019-05-01', '2019-08-20']
data_type_to_plot = 'grid'
exclude_days_to_plot = []
avg_to_plot = 'avghour'
# home_id_to_plot = int(home_id_selector.value)
# print(attr, old, new)
# if home_id_selector.value == new:
# daterange_to_plot = [all_min_date[home_id_to_plot].strftime("%Y-%m-%d"), all_max_date[home_id_to_plot].strftime("%Y-%m-%d")]
# date_range_slider.start=all_min_date[home_id_to_plot]
# date_range_slider.end=all_max_date[home_id_to_plot]
# date_range_slider.value=(all_min_date[home_id_to_plot],all_max_date[home_id_to_plot])
# print("if 1 truth")
# else:
daterange_raw = list(date_range_slider.value_as_datetime)
daterange_to_plot = [daterange_raw[0].strftime("%Y-%m-%d"), daterange_raw[1].strftime("%Y-%m-%d")]
# print("falsyy")
# date_range_slider.update(start=all_min_date[home_id_to_plot], end=all_max_date[home_id_to_plot],
# value=(all_min_date[home_id_to_plot], all_max_date[home_id_to_plot]))
# print("after:",new)
# print(attr,old,new)
# daterange_to_plot=list(string_date1,string_date2)
# daterange_to_plot=dummy_daterange
data_type_to_plot = "grid" # data_type_available[data_type_selector.active]
# only working button call so far
granularity_to_plot = granularity_1.labels[granularity_1.active]
# granularity_to_plot= [granularity_1.labels[i] for i in granularity_1.active]
print("before data create")
# create a new data source
new_src1 = plot1_data(home_id_to_plot, daterange=daterange_to_plot, data=data_type_to_plot,
xaxis=granularity_to_plot)
new_src2 = plot2_data(home_id_to_plot, daterange=daterange_to_plot,
weekdays=exclude_days_to_plot, data=data_type_to_plot,xaxis=avg_to_plot)
print("before data update")
# push new data to the source data the rest of the app is usig for plot1
src1.data.update(new_src1.data)
src2.data.update(new_src2.data)
print("updated data")
# def new_home(attr,old,new):
# return
############# Add widgets
# only the granularity implemented so far
granularity_1 = RadioGroup(
labels=["15 Minutes", "Hour", "Day", "Week", "Month"], active=0)
granularity_1.on_change('active',
update) # not sure exactly how this works but runs update on the change of the button and passes through the value of the button
home_ids_available = np.unique(filterData['dataid'])
home_ids_available= list(map(str, home_ids_available))
home_id_selector = Dropdown(label="Home ID to Plot", button_type="warning", menu=home_ids_available)
home_ids_available = list(map(str, home_ids_available))
home_id_selector = Dropdown(label="Home ID to Plot", button_type="warning", menu=home_ids_available, value="27")
# home_id_selector.on_change('value', update) #############put back in later!!!!1
date_range_slider = DateRangeSlider(title="Date Range: ", start=date(2019, 5, 1), end=date(2019, 8, 20),
value=(date(2019, 5, 1), date(2019, 8, 20)), step=1, callback_policy = 'mouseup')
date_range_slider.on_change("value", update)
# data_type_selector = RadioButtonGroup(labels=["Net Home Consumption","Solar Generation","EV Consumption"],active=0)
# data_type_selector.on_change('active',update)
# data_type_available=["grid","solar","car1"]
############ Initialize opening plot and data
src1 = plot1_data(int(home_ids_available[0]), ['2019-05-01', '2019-08-20'], 'grid',
'15 Minutes') # start with a data range we know is correct
plot1 = plot1_plot(src1)
src2 = plot2_data(27,['2019-05-01', '2019-08-20'],[],'grid','avghour')
plot2 = plot2_plot(src2)
##### Formatting of the app screen
# Put controls in a single element (add more later to format)
controls = WidgetBox(granularity_1, home_id_selector, date_range_slider) # data_type_selector)
# Create a row layout
layout = row(controls, column(plot1,plot2))
# Make a tab with the layout
tab = Panel(child=layout, title='First Tab')
return tab
def create_tab(data, name):
def make_dataset(metric_fun, metric_sentiment, month_start, month_end,
editorial, category, social_network, product):
"""Constrói os datasets para cada tipo de gráfico utilizado no dashboard
Parâmetros
----------
data : DataFrame
Pandas DataFrame expandido com dados do Sprinklr
metric_fun : FUN
Função para calcular métrica específica selecionada no widget
metric_sentiment : str
Sentimento relacionado à métrica escolhida (Positividade: positivo, Gradiente: negativo, Crise: negativo, Saúde do post: positivo)
[restante] : str
Valaores selecionados nas opções de filtros nos widgets
Retorna
-------
dict
Dicionário com três chaves, correspondentes aos três gráficos apresentados. Cada chave é relacionada ao nome o gráfico
e os valores são datasets no formato column data source
"""
month_start = pd.Timestamp(month_start)
month_end = pd.Timestamp(month_end)
# Filtragem dos dados com base nas seleções dos widgets
filters = {
'Editoria': editorial,
'Categoria': category,
'SocialNetwork': social_network,
'Produto': product
}
filtered_data = filter_data(data, filters)
# Gera datasets para cada gráfico
ts_data = metric_fun(filtered_data)
ts_data = ts_data[(ts_data.time >= month_start)
& (ts_data.time <= month_end)]
donut_data = filtered_data[(filtered_data.Month >= month_start)
& (filtered_data.Month <= month_end)]
donut_data = percent(donut_data)
donut_data['angle'] = donut_data['value'] / sum(
donut_data['value']) * 2 * pi
donut_data['color'] = Category20c[donut_data.shape[0]]
avg_donut_data = percent_avg(filtered_data)
avg_donut_data = avg_donut_data[avg_donut_data.Month == month_end][[
'Sentimento', 'MAVG'
]]
avg_donut_data.columns = ['label', 'value']
avg_donut_data['angle'] = avg_donut_data['value'] / sum(
avg_donut_data['value']) * 2 * pi
avg_donut_data['color'] = Category20c[avg_donut_data.shape[0]]
top_data = filter_sentiment(filtered_data, metric_sentiment)
top_data = top_data[(top_data.Month >= month_start)
& (top_data.Month <= month_end)]
top_data = brand_health_txengages(top_data)
avg_top_data = round(top_data.score.mean(), 2)
top_data = top_data.sort_values('score', ascending=False).iloc[:10]
top_data = top_data.sort_values('score')
top_data['recorte'] = [
'1', '2', '3', '4', '5', '6', '7', '8', '9', '10'
]
# Converte dataframes em column data source
datasets = {
'ts': ColumnDataSource(ts_data),
'donut': ColumnDataSource(donut_data),
'avg_donut': ColumnDataSource(avg_donut_data),
'top': ColumnDataSource(top_data),
'avg_top': avg_top_data
}
return datasets
def update(attr, old, new):
"""Constrói os datasets para cada tipo de gráfico utilizado no dashboard
Parâmetros
----------
old : ColumnDataSource
Dataframe antigo relacionado aos filtros antigos
new : ColumnDataSource
Dataframe novo, com linhas filtradas de acordo com seleções mais recentes
"""
month_start = month_select.value_as_date[0]
month_end = month_select.value_as_date[1]
editorial = editorial_select.value
category = category_select.value
product = product_select.value
social_network = [
social_network_select.labels[i]
for i in social_network_select.active
]
metric = metric_select.value
metric_attr = get_metric_attr(metric)
metric_fun = metric_attr['fun']
metric_sentiment = metric_attr['sentiment']
new_src = make_dataset(metric_fun=metric_fun,
metric_sentiment=metric_sentiment,
month_start=month_start,
month_end=month_end,
editorial=editorial,
category=category,
social_network=social_network,
product=product)
src['ts'].data.update(new_src['ts'].data)
src['top'].data.update(new_src['top'].data)
src['avg_top'] = new_src['avg_top']
src['donut'].data.update(new_src['donut'].data)
src['avg_donut'].data.update(new_src['avg_donut'].data)
networks = data.SocialNetwork.unique().tolist()
editorials = get_multselect_options(data, 'Editoria')
categories = get_multselect_options(data, 'Categoria')
products = get_multselect_options(data, 'Produto')
month_select = DateRangeSlider(start=date(2019, 1, 1),
end=date(2019, 8, 1),
value=(date(2019, 1, 1), date(2019, 8, 1)),
step=1,
format="%b %Y")
metric_select = Select(value="gradiente",
options=[("velocity", "Parâmetro de Crise"),
("positivity", "Grau de Positividade"),
("gradiente", "Grau de Negatividade"),
("brand_health", "Saúde da Marca"),
("post_health", "Saúde do Post")])
product_select = MultiSelect(value=['Todos'], options=products)
category_select = MultiSelect(value=['Todos'], options=categories)
editorial_select = MultiSelect(value=['Todos'], options=editorials)
social_network_select = CheckboxGroup(labels=networks,
active=list(range(len(networks))))
metric_select.on_change('value', update)
month_select.on_change('value', update)
editorial_select.on_change('value', update)
category_select.on_change('value', update)
product_select.on_change('value', update)
social_network_select.on_change('active', update)
initial_metric_attr = get_metric_attr(metric_select.value)
metric_sentiment = initial_metric_attr['sentiment']
initial_networks = [
social_network_select.labels[i] for i in social_network_select.active
]
src = make_dataset(metric_fun=initial_metric_attr['fun'],
metric_sentiment=metric_sentiment,
month_start=month_select.value_as_date[0],
month_end=month_select.value_as_date[1],
editorial=editorial_select.value,
category=category_select.value,
social_network=initial_networks,
product=product_select.value)
p_ts = make_plot_ts(src['ts'], 'Evolução', metric_sentiment)
p_top = make_dotplot(src['top'])
avg_top = src['avg_top']
avg_top = create_div_title(f'Escore Médio: {avg_top}')
p_donut = make_plot_donut(src['donut'], 'Percentual')
p_avg_donut = make_plot_donut(src['avg_donut'], 'Norma Percentual')
metric_title = create_div_title('MÉTRICA')
month_title = create_div_title('PERÍODO')
network_title = create_div_title('REDE SOCIAL')
category_title = create_div_title('CATEGORIA')
editorial_title = create_div_title('EDITORIA')
product_title = create_div_title('PRODUTO')
controls = WidgetBox(
column(metric_title,
metric_select,
Div(height=5),
month_title,
month_select,
Div(height=5),
editorial_title,
editorial_select,
Div(height=5),
category_title,
category_select,
Div(height=5),
product_title,
product_select,
Div(height=5),
network_title,
social_network_select,
width=250))
plots = column(p_ts, Div(height=20), row(p_donut, p_avg_donut))
layout = row(controls, Div(width=50), plots)
layout = column(Div(text="", height=5), layout, Div(width=20), avg_top,
p_top)
tab = Panel(child=layout, title=name)
return tab
pdict['active'] = radio_button_group.active
completed.text = '<b>Update Completed!</b>'
sdate_input = TextInput(value="2020-01-03 00:00:00",
title="Start Date: (YYYY-MM-DD HH:MM:SS)")
#sdate_input.on_change("value", my_slider_handler)
edate_input = TextInput(value="2020-01-03 01:00:00",
title="End Date: (YYYY-MM-DD HH:MM:SS)")
#edate_input.on_change("value", my_slider_handler)
date_range_slider = DateRangeSlider(title="Data Filter Date Range: ",
start=datetime.datetime(2020, 1, 3, 0),
end=datetime.datetime(2020, 3, 20, 1),
value=(datetime.datetime(2020, 1, 3, 0),
datetime.datetime(2020, 1, 3, 1)),
format="%x,%X")
date_range_start = TextInput(
value="2020-01-03 00:00:00",
title="Data Filter Start Date: (YYYY-MM-DD HH:MM:SS)")
#sdate_input.on_change("value", my_slider_handler)
date_range_end = TextInput(value="2020-01-03 01:00:00",
title="Data Filter End Date: (YYYY-MM-DD HH:MM:SS)")
idle_range_slider = DateRangeSlider(title="Idle Date Range: ",
start=datetime.datetime(2020, 1, 3, 0),
end=datetime.datetime(2020, 3, 20, 1),
value=(datetime.datetime(2020, 1, 3, 0),
source_wed = ColumnDataSource(data=dict(time=[], y_value=[]))
source_thu = ColumnDataSource(data=dict(time=[], y_value=[]))
source_fri = ColumnDataSource(data=dict(time=[], y_value=[]))
source_sat = ColumnDataSource(data=dict(time=[], y_value=[]))
source_sun = ColumnDataSource(data=dict(time=[], y_value=[]))
# Create Input controls
station_name = Select(title="Select station",
value="Varsapuistikko",
options=stations_names)
y_axis = Select(title="Y Axis",
options=sorted(axis_map.keys()),
value="empty_slots")
date_range_slider = DateRangeSlider(title="Date Range: ",
start=min_date,
end=max_date,
value=(min_date, max_date),
step=1)
p = figure(x_axis_type="datetime",
plot_height=300,
plot_width=500,
title="",
tools="",
toolbar_location=None)
p_mon = figure(plot_height=200,
plot_width=400,
toolbar_location=None,
tools="")
p_tue = figure(plot_height=200,
plot_width=400,
#sdate_input.on_change("value", my_slider_handler)
edate_input = TextInput(value="2020-01-03 01:00:00",
title="End Date: (YYYY-MM-DD HH:MM:SS)")
#edate_input.on_change("value", my_slider_handler)
count_threshold = TextInput(value="1", title="Count Threshold Value")
resolution_slider = Slider(start=1,
end=15,
value=7,
step=1,
title="Hex Resolution")
sdate_range_slider = DateRangeSlider(title="Date Range: ",
start=datetime.datetime(2019, 8, 15, 0),
end=datetime.datetime(2020, 4, 28, 1),
value=(datetime.datetime(2020, 1, 3, 1),
datetime.datetime(2020, 1, 3, 2)),
format="%x,%X")
#sdate_range_slider = DateRangeSlider(title="Date Range: ", start=datetime.datetime(2017, 1, 1,1), end=datetime.datetime(2017, 2, 7,2), value=(datetime.datetime(2017, 9, 7,1), datetime.datetime(2017, 9, 7,2)),format="%Y-%m-%d %H")
#sdate_range_slider = DateRangeSlider(title="Date Range: ", start=datetime.datetime(2017, 1, 1,1), end=datetime.datetime(2017, 2, 7,2), value=(datetime.datetime(2017, 9, 7,1), datetime.datetime(2017, 9, 7,2)),step=1)
#sdate_range_slider.on_change("value", my_slider_handler)
#sdate_range_slider = DateSlider(title="Date Range: ", start=datetime.datetime(2017, 1, 1,1), end=datetime.datetime(2019, 9, 7,2), value=(datetime.datetime(2017, 9, 7,1), datetime.datetime(2017, 9, 7,2)), step=1)
sMindate = datetime.datetime(2020, 1, 3, 0)
Mindate = datetime.datetime(2019, 8, 15, 0)
date_text = Div(text='<b style="color:black">' + str(Mindate) +
' to ' +
str(Mindate + timedelta(hours=6)) + '<br></b>',
width=100,
height=50,
orientation='vertical',
button_type='primary')
save_button.on_click(save_png_callback)
grid = gridplot(children=[
word_barplots,
],
toolbar_location=None,
merge_tools=True)
date_range_slider = DateRangeSlider(
start=data['time'].iloc[0],
end=data['time'].iloc[-1],
value=(data['time'].iloc[0], data['time'].iloc[-1]),
format='%d/%m@%H:%M',
step=1,
width=95 * (len(word_barplots) - 1) - 40, # padded
bar_color='purple')
play_button = Button(label='Run', width=75, button_type='success')
play_button.on_click(update)
bottom_layout = Row(children=[
date_range_slider,
play_button,
])
svg_layout = Row(svg_div, width=450, height=380)
arroba_layout = Row(children=[
def select_values_2(attr, old, new, w_box, c_data):
if new != 'None':
if c_data[w_box.children[1].
value][new].values.dtype == 'object': # categorical data
level_3 = MultiSelect(title='value',
value=['None'],
options=['None'],
width=180)
try:
level_3.options = np.unique(
c_data[w_box.children[1].value]
[new].iloc[:, 0].dropna().values).tolist()
level_3.value = [level_3.options[0]]
except TypeError:
level_3.options = np.unique([
str(obj) for obj in c_data[w_box.children[1].value]
[new].iloc[:, 0].dropna().values
]).tolist()
finally:
w_box.children[3] = column(level_3)
elif 'datetime' in str(c_data[w_box.children[1].value]
[new].values.dtype): # datetime data
start = c_data[w_box.children[1].value][new].min().dt.date.item()
end = c_data[w_box.children[1].value][new].max().dt.date.item()
date_slider = DateRangeSlider(
title="",
start=start,
end=end,
value=(start, end),
# value_as_date=True,
# step=1,
width=180)
checkbox_group = CheckboxGroup(labels=["invert selection"],
active=[],
width=180)
w_box.children[3] = column(date_slider, checkbox_group)
elif 'int' in str(c_data[w_box.children[1].value][new].values.dtype) or \
'float' in str(c_data[w_box.children[1].value][new].values.dtype):
# print("3 ", clinical_data[select_1.value][new].values.dtype)
start = c_data[w_box.children[1].value][new].min().item()
end = c_data[w_box.children[1].value][new].max().item()
slider = RangeSlider(start=start,
end=end,
step=0.1,
value=(start, end),
title=new + " Range",
width=180)
checkbox_group = CheckboxGroup(labels=["invert selection"],
active=[],
width=180)
w_box.children[3] = column(slider, checkbox_group)
else:
print(
"Something went wrong, unexpected datatype by clinical data value selecting"
) # TODO error message?
else:
w_box.children[3] = PreText(text='please select a property', width=200)
Div
Elemento do tipo Div do Bokeh
"""
text = f'<b><font size=2 color="#00004C">{title}</font></b>'
div = Div(text=text)
return div
networks = data.SocialNetwork.unique().tolist()
editorials = get_multselect_options('Editoria')
categories = get_multselect_options('Categoria')
products = get_multselect_options('Produto')
month_select = DateRangeSlider(start=date(2019, 1, 1),
end=date(2019, 9, 1),
value=(date(2019, 1, 1), date(2019, 9, 1)),
step=1,
format="%b %Y")
metric_select = Select(title="Métrica:",
value="velocity",
options=[("velocity", "Parâmetro de Crise"),
("positivity", "Grau de Positividade"),
("gradiente", "Grau de Negatividade"),
("brand_health", "Saúde da Maca"),
("post_health", "Saúde do Post")])
product_select = MultiSelect(value=['Todos'], options=products)
category_select = MultiSelect(value=['Todos'], options=categories)
editorial_select = MultiSelect(value=['Todos'], options=editorials)
social_network_select = CheckboxGroup(labels=networks,
def build(self):
# init widget
if self.has_type == True:
active = [i for i in range(len(self.labels))]
labels = [
self.labels[i] + " (" + self.colors[i] + ")"
for i in range(len(self.labels))
]
self.label_checkbox = CheckboxGroup(labels=labels, active=active)
self.x_rangeslider = RangeSlider(value=[-100, 150],
start=-100,
end=150,
step=1,
title="x")
self.y_rangeslider = RangeSlider(value=[-100, 100],
start=-100,
end=100,
step=1,
title="y")
if self.has_time == True:
self.time_rangeslider = DateRangeSlider(
format="%Y-%b-%d %H:%M",
value=[self.timeRange[0], self.timeRange[1]],
start=self.timeRange[0],
end=self.timeRange[1],
step=10,
title="time")
controls = []
if self.has_type == True:
controls.append(self.label_checkbox)
controls.append(self.x_rangeslider)
controls.append(self.y_rangeslider)
if self.has_time == True:
controls.append(self.time_rangeslider)
widgets = widgetbox(controls, width=700, height=700)
# if self.has_type == True:
# widgets = widgetbox(self.label_checkbox, self.x_rangeslider, self.y_rangeslider, width=700, height=700)
# controls = [self.label_checkbox, self.x_rangeslider, self.y_rangeslider]
# else:
# widgets = widgetbox(self.x_rangeslider, self.y_rangeslider, width=700, height=700)
# controls = [self.x_rangeslider, self.y_rangeslider]
for control in controls:
if type(control) == CheckboxGroup:
control.on_change('active',
lambda attr, old, new: self.update())
else:
control.on_change('value',
lambda attr, old, new: self.update())
# self.source_scatter is variable which figure will display the data from
self.source_scatter = ColumnDataSource(data=self.original_data)
self.scatter_plot = figure(plot_width=700,
plot_height=700,
tools=self.tools,
tooltips=self.tooltip,
toolbar_location="left",
title="")
sp = self.scatter_plot.circle(x='x',
y='y',
size=20,
color=self.color,
source=self.source_scatter)
table_attr_name = self.attr_name[:-1]
st_table = selection_table(table_attr_name, sp, self.source_scatter)
layout = column(row(widgets, self.scatter_plot), st_table.table)
self.update()
for name in self.attr_name:
self.source_scatter.data[name] = pd.Series(
self.source_scatter.data[name])
curdoc().add_root(layout)
states_select2.on_change('active', update)
states_select3 = CheckboxGroup(labels=STATES[36:54], active=[])
states_select3.on_change('active', update)
# states_selection = CheckboxGroup(labels=STATES, active = [0,1])
# states_selection.on_change('active', update)
thresh_select = Slider(start=0,
end=1000,
step=1,
value=0,
title='Case Count Minimum')
thresh_select.on_change('value', update)
range_select = DateRangeSlider(start=dt.date(2020, 1, 21),
end=dt.date.today(),
value=(dt.date(2020, 1, 21), dt.date.today()),
step=1,
title='Date Range')
range_select.on_change('value', update)
select_all = Button(label="select all")
select_all.on_click(activate_all_update)
unselect_all = Button(label="unselect all")
unselect_all.on_click(deactivate_all_update)
# Initialize source
initial_states = [states_select1.labels[i] for i in states_select1.active] + \
[states_select2.labels[i] for i in states_select2.active] + \
[states_select3.labels[i] for i in states_select3.active]
src = make_dataset(initial_states,
labels = list(labels)
values = list(values)
for i in all_gate_names:
if i not in labels:
labels.append(i)
values.append(0)
zipped = sorted(zip(labels, values))
labels, values = zip(*zipped)
labels = np.asarray(labels)
values = np.asarray(values)
source.add(values, name=car_type)
date_slider = DateRangeSlider(title='Date Range',
start=date(2015, 5, 1),
end=date(2016, 5, 30),
value=(date(int(start_y), int(start_m),
int(start_d)),
date(int(end_y), int(end_m), int(end_d))),
step=1)
color = ['blue', 'red', 'yellow', 'purple', 'cyan', 'black', 'green']
def date_range_update(attrname, old, new):
"""
Callback for Date Range Slider Interaction.
All of previous created for the ColumnDataSource
must be redone in the update in order to correctly
interact with the graph.
"""
d1 = datetime.fromtimestamp(date_slider.value[0] / 1000)
def timeseries_tab(dataframe):
"""
return a tab showing steps vs time for each person
"""
def make_dataset(name_list,
range_start='2019-01-01',
range_end='2019-12-31'):
"""
Filter the full dataset by name and by date range,
and return it as a Bokeh ColumnDataSource
"""
## why do I have to do this? What is the point of a DateRangeSlider???
if isinstance(range_start, int):
range_start = datetime.fromtimestamp(range_start / 1000)
if isinstance(range_end, int):
range_end = datetime.fromtimestamp(range_end / 1000)
# filtered_df = dataframe[name_list].loc[range_start: range_end]
filtered_df = dataframe.loc[range_start:range_end]
source = ColumnDataSource(filtered_df)
return source
def style(p):
# Title
p.title.align = 'center'
p.title.text_font_size = '20pt'
p.title.text_font = 'serif'
# Axis titles
p.xaxis.axis_label_text_font_size = '14pt'
p.xaxis.axis_label_text_font_style = 'bold'
p.yaxis.axis_label_text_font_size = '14pt'
p.yaxis.axis_label_text_font_style = 'bold'
# Tick labels
p.xaxis.major_label_text_font_size = '12pt'
p.yaxis.major_label_text_font_size = '12pt'
return p
def make_plot(source):
"""
create a Bokeh figure with the selected data
"""
names_to_plot = source.column_names[1:]
time_series = figure(x_axis_type="datetime",
plot_width=700,
plot_height=550)
hover = HoverTool(tooltips=[("Name", "$name"), ("Date", "@Date{%F}"),
("Steps", "$y")],
formatters={'Date': 'datetime'})
time_series = style(time_series)
time_series.add_tools(hover)
time_series.legend.location = "top_left"
for i, name in enumerate(names_to_plot):
time_series.line("Date",
name,
source=source,
line_color=Category20_16[i],
line_width=2,
name=name,
legend_label=name)
return time_series
def update(attr, old, new):
"""
Update data source when something is changed.
"""
name_list = [name_selection.labels[i] for i in name_selection.active]
new_src = make_dataset(name_list, date_slider.value[0],
date_slider.value[1])
cds.data.update(new_src.data)
def update_lines(attr, old, new):
"""
Hide selected lines
"""
names_to_plot = [
name_selection.labels[i] for i in name_selection.active
]
for name in names:
if name in names_to_plot:
p.select_one({"name": name}).visible = True
else:
p.select_one({"name": name}).visible = False
### back to the timeseries_tab function
names = list(dataframe.columns)
names.sort()
# widgets to allow user to configure the plot
name_selection = CheckboxGroup(labels=names,
active=list(range(len(names))))
name_selection.on_change('active', update_lines)
date_slider = DateRangeSlider(title="Date range",
start=date(2019, 1, 1),
end=date(2019, 12, 31),
value=(date(2019, 1, 1), date(2019, 12, 31)),
step=1)
date_slider.on_change('value', update)
initial_names = [name_selection.labels[i] for i in name_selection.active]
cds = make_dataset(initial_names, date_slider.value[0],
date_slider.value[1])
p = make_plot(cds)
controls = WidgetBox(name_selection, date_slider)
layout = row(controls, p)
tab = Panel(child=layout, title="Time series")
return tab
source.data = source.from_df(df)
source_static.data = source.data
one_day_df = get_data_subset((VARS['selected_day'], VARS['selected_day']))
single_day_source.data = single_day_source.from_df(one_day_df)
#############################
# Initialize interactive widgets before calling first update
#############################
update()
date_selector = DateRangeSlider(title="Select Timeframe",
end=VARS['global_end'],
start=VARS['global_start'],
step=1,
value=(
VARS['time_range'][0],
VARS['time_range'][1],
))
single_day_selector = DateSlider(
title="Select Day",
end=VARS['global_end'],
start=VARS['global_start'],
step=1,
value=VARS['selected_day'],
)
date_selector.on_change('value', update_time_range)
single_day_selector.on_change('value', update_highlighted_day)
##################################################################
range_slider = RangeSlider(title="Numerical range",
value=[30, 70],
start=0,
end=100,
step=0.5)
date_slider = DateSlider(title="Date",
value=date(2014, 1, 1),
start=date(2010, 1, 1),
end=date(2020, 1, 1),
step=1)
date_range_slider = DateRangeSlider(title="Date range",
value=(date(2014, 1,
1), date(2018, 12, 31)),
start=date(2010, 1, 1),
end=date(2020, 1, 1),
step=1)
only_value_slider = Slider(value=50, start=0, end=96, step=5)
no_title_slider = Slider(title=None, value=50, start=0, end=96, step=5)
def color_picker():
def color_slider(title, color):
return Slider(title=title,
show_value=False,
value=127,
start=0,
end=255,
p1.line('x', 'y2', source = source, legend="KOSPI200", color = 'grey',alpha = 0.8,
muted_color = 'grey', muted_alpha = 0.2, line_width=2)
p1.legend.location = "top_left"
p1.legend.click_policy="mute"
p1.add_tools(HoverTool(tooltips=[('Date','@x{%Y-%m-%d}'), ('Port','@y1'), ('BM', '@y2')],
formatters={'x' : 'datetime'},
mode = 'vline',
show_arrow=False, point_policy='follow_mouse'))
p2 = figure(plot_width=1200, plot_height=200, x_range = p1.x_range, x_axis_type="datetime")
p2.varea(x='x', y1 = 'zeros', y2 = 'dd', source = source)
date_slider = DateRangeSlider(title="Date Range: ",
start=ts.index[0],
end=ts.index[-1],
value=(ts.index[0], ts.index[-1]),
step=1)
def update_data(attrname, old, new):
# Get the current slider values
x_start = datetime.fromtimestamp(date_slider.value[0]/1000)
x_end = datetime.fromtimestamp(date_slider.value[1]/1000)
x_start = pd.to_datetime(x_start)
x_end = pd.to_datetime(x_end)
#print(x_start)
#print(x_end)
# Generate new data
new_df = df[(df['x']>= x_start) & (df['x'] <= x_end)]
new_df.loc[:,'y1'] = (new_df['y1'].pct_change().fillna(0) + 1).cumprod() * 100
new_df.loc[:,'y2'] = (new_df['y2'].pct_change().fillna(0) + 1).cumprod() * 100
