WorkPlotly.py

# -*- coding: utf-8 -*-
"""
Create interactive line plots using Plotly.

Script is now obsolete, but preserved for remembering my work on it.
"""

import numpy as np

import plotly
import plotly.express as px
from plotly.offline import plot as pplot
import plotly.graph_objects as go

import covid


#%% Get the data

state = covid.read_covid_data_wi(dataset='state')


#%% Manipulate data columns for prep

# reduce and rename columns
col_rename = {'Date':'Date','POS_NEW': 'Cases', 'TEST_NEW': 'Tests', 'DTH_NEW': 'Deaths', 'HOSP_NEW': 'Hospitalizations'}
col_cumul = {'POSITIVE': 'Total Cases', 'NEGATIVE': 'Total Negative', 'HOSP_YES': 'Total Hospitalizations', 'DEATHS': 'Total Deaths'}
state = state[col_rename.keys()]
state = state.rename(columns=col_rename)

#%% Standard plots
plotpath = '.\\docs\\assets\\plotly'

# Cases / Tests
covid.plotly_casetest(sourcedata=state, 
                      case_col='Cases', 
                      test_col='Tests', 
                      date_col='Date', 
                      savefile=plotpath + '\\Cases-Tests-WI.html',
                      )

# Deaths / Hospitalizations
covid.plotly_deadhosp(sourcedata=state, 
                      hosp_col='Hospitalizations', 
                      dead_col='Deaths', 
                      date_col='Date', 
                      savefile=plotpath + '\\Deaths-Hosp-WI.html'
                      )


#%% Old parts of script
if False:
    #%%
    
    # # Deaths / Hospitalizations - using twolines explicitly
    # covid.plotly_twolines(
    #     sourcedata=state,
    #     column1='Deaths',
    #     column2='Hospitalizations',
    #     date_col='Date',
    #     plotcolors=['firebrick', 'darkorange', 'rosybrown'],
    #     savefile=plotpath + '\\Deaths-Hosp-WI-temp.html',
    #     column1_bar=True,
    #     plotlabels=dict(title='WI Daily Deaths and Hospitalizations',
    #                     yaxis='Deaths / Hospitalizations',
    #                     ),
    #     )
    #%% Hospitalizations / Deaths
    
    hosp_death_factor = 1
    
    range_max = max(state_avg.Deaths.max(), state_avg.Hospitalizations.max()/hosp_death_factor)
    range_deaths = np.array([-range_max * 0.05, 1.05*range_max])
    
    fig = plotly.subplots.make_subplots()
    
    # individual deaths bar chart
    fig.add_trace(
        go.Bar(x=state.index, 
               y=state.Deaths,
               name='Deaths', 
               marker_color='rosybrown', 
               hovertemplate='%{y:.0f}'),)
    
    # death 7-day avg
    fig.add_trace(
        go.Scatter(x=state_avg.index, 
               y=state_avg.Deaths,
               name='Deaths (7-day avg)', 
               line_color='firebrick', 
               hovertemplate='%{y:.1f}'),)
    
    # new hospitalization 7-day average
    fig.add_trace(
        go.Scatter(x=state_avg.index, 
                   y=state_avg.Hospitalizations, 
                   name='Hospitalizations (7-day avg)', 
                   line_color='darkorange', 
                   hovertemplate='%{y:.1f}'),)
    
    
    fig.update_yaxes({'range': range_deaths}, secondary_y=False, title_text='Daily deaths / hospitalizations')
    fig.update_yaxes({'range': range_deaths*hosp_death_factor}, secondary_y=True, title_text='Daily hospitalizations')
    fig.update_layout(title_text='WI Daily Deaths and Hospitalizations',
                      hovermode='x unified',
                      legend=dict(yanchor="top", y=0.99, xanchor="left", x=0.01))
    
    # fig.add_annotation(dict(text='*Deaths are daily reported values, Hospitalizations are 7-day averages.',
    #                         x=0.05, y=-0.1, 
    #                         xref='paper', yref='paper',
    #                         showarrow=False))
    
    # plot and save as html, with plotly JS library in separate file in same directory
    pplot(fig, 
          filename = plotpath + '\\Deaths-Hosp-WI.html',
          include_plotlyjs='cdn')
    
    
    
    #%% OLD: Cases/Tests
    
    # compute y axis range
    # want tests to be on a scale exactly 10x cases
    range_max = max(state_avg.Tests.max()/10, state_avg.Cases.max())
    range_cases = np.array([-range_max * 0.05, 1.05*range_max])
    
    
    fig = plotly.subplots.make_subplots(specs=[[{"secondary_y": True}]])
    
    # # individual cases bar chart
    # fig.add_trace(
    #     go.Bar(x=state.index, 
    #            y=state.Cases,
    #            name='Cases', 
    #            marker_color='lightsteelblue', 
    #            hovertemplate='%{y:.0f}'),)
    
    # # individual tests bar chart
    # fig.add_trace(
    #     go.Bar(x=state.index, 
    #            y=state.Tests,
    #            name='Tests', 
    #            marker_color='darkkhaki', 
    #            hovertemplate='%{y:.0f}'),
    #     secondary_y=True)
    
    
    fig.add_trace(
        go.Scatter(x=state_avg.index, 
                   y=state_avg.Cases, 
                   name='Cases (7-day avg)', 
                   line_color='steelblue', 
                   hovertemplate='%{y:.0f}'),
        secondary_y=False)
    
    fig.add_trace(
        go.Scatter(x=state_avg.index, 
                   y=state_avg.Tests, 
                   name='Tests (7-day avg)', 
                   line_color='olivedrab', #darkolivegreen
                   hovertemplate='%{y:.0f}'),
        secondary_y=True)
    
    
    fig.update_yaxes({'range': range_cases}, secondary_y=False, title_text='Daily cases')
    fig.update_yaxes({'range': range_cases*10}, secondary_y=True, title_text='Daily new people tested')
    fig.update_layout(title_text='WI Daily Cases and Tests',
                      hovermode='x unified',
                      legend=dict(yanchor="top", y=0.99, xanchor="left", x=0.01))                
                    
    # plot and save as html, with plotly JS library in separate file in same directory
    pplot(fig, 
          filename = plotpath + '\\Cases-Tests-WI.html', 
          include_plotlyjs='cdn')