COVID-19 Wikipedia pageview spikes, 2019-2022 (plotly functions)

file.py

dfToPlot = allThePandasFiltered.copy()

def do_pyplot( data, title, logY=False, unstackAndSort=False, grid=False, trends=False, plotMax=False, ylim=False ):
    pd.options.plotting.backend = "matplotlib"

    fig,ax = plt.subplots(figsize=plotSize)
    plt.title(title)
    ax.set_xlabel("Date")
    ax.set_ylabel("Pageviews")
    if ylim:
        ax.set_ylim(ylim)
    if(logY):
        ax.set_yscale('log')
        ax.set_ylabel("Pageviews (log)")
    ax.get_yaxis().set_major_formatter(
        mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
    
    ax.grid(grid)
    
    if unstackAndSort:
        data = data.unstack()
        data = data[data.max().sort_values(ascending=True).index]
        
        # Only get the top ones if requested
        if unstackAndSort > 1:
            data = data.iloc[:,-unstackAndSort:]
        
        data.plot.area(ax=ax, stacked=True)
        
        handles, labels = ax.get_legend_handles_labels()
        ax.legend(reversed(handles), reversed(labels), title=None)
        ax.get_yaxis().set_major_formatter(
            mtick.FuncFormatter(lambda x, p: format(int(x), ',')))
    else:
        data.plot(label='daily')

        if trends:
            for index in trendLines:
                data.rolling(window=index).mean().plot(label=trendLines[index]['label'], linestyle=trendLines[index]['style'])

        ax.legend()

    if plotMax:
        maxIndex, maxValue = data.idxmax(), data.max()
        ax.annotate(format(int(maxValue), ','), xy=(maxIndex, maxValue), xytext=(maxIndex, maxValue+(maxValue/100)))

    return fig, ax

def add_plotly_common(fig):
    # https://plotly.com/python-api-reference/generated/plotly.graph_objects.Figure.html?highlight=add_vline#plotly.graph_objects.Figure.add_vline
    # Got to use vrect, not line because of https://community.plotly.com/t/error-if-adding-annotation-to-add-vline-on-plotly-express-when-x-is-a-datetime/52497
    for event in events:
        fig.add_vrect(line_width=1, line_dash="dash", line_color="lightgrey", annotation_textangle=-90, annotation_font=dict( color="grey", size=12), x0=event['date'], x1=event['date'], annotation_text=event['text'])

    fig.update_layout(dict(updatemenus=[
                        dict(
                            type = "buttons",
                            direction = "left",
                            buttons=list([
                                # https://stackoverflow.com/a/69847069/4746236
                                dict(
                                    args=["visible", "legendonly"],
                                    label="Deselect All",
                                    method="restyle"
                                ),
                                dict(
                                    args=["visible", True],
                                    label="Select All",
                                    method="restyle"
                                ),
                                # https://stackoverflow.com/a/61237669/4746236
                                dict(
                                    args=[{'yaxis.type': 'linear'}],
                                    label="Linear Scale",
                                    method="relayout"
                                ),
                                dict(
                                    args=[{'yaxis.type': 'log'}],
                                    label="Log Scale",
                                    method="relayout"
                                ),
                            ]),
                            pad={"r": 10, "t": 10},
                            showactive=False,
                            x=1,
                            xanchor="right",
                            y=1.1,
                            yanchor="top"
                        ),
                    ]
              ))

    fig.update_layout(
        showlegend=True,
        width=2000,
        height=1000,
    )

def add_plotly_title(fig, title):
    genText = "Using current and previous page titles linked to Wikidata items in any of the following queries(" + ", ".join(lookupWikidataStatements) + ")"
    pageviewText = "and the Wikimedia page view data https://meta.wikimedia.org/wiki/Research:Page_view (data points less than 1k views removed). By @addshore"
    
    fig.update_layout(
        title=go.layout.Title(
            text=title+"<br><sup>"+genText+"<br>"+pageviewText + "</sup>",
            xref="paper",
            x=0
        ),
    )
    
def do_plotly_simple(data, title):
    pd.options.plotting.backend = "plotly"
    fig = data.plot(
        labels=dict(index="date",value="pageviews",variable=""),
    )
    add_plotly_common(fig)
    add_plotly_title(fig,title)
    show_and_save_plotly_fig(fig,title)

def do_plotly_group(rawData, groupBy, title):
    grouped=rawData.groupby(['date',groupBy])['pageviews'].sum().reset_index()
    grouped = grouped.sort_values(by=['date', 'pageviews']).reset_index(drop=True)
    group = grouped.groupby(groupBy)

    fig = go.Figure()

    # each group iteration returns a tuple
    # (group name, dataframe)
    for group_name, df in group:
        fig.add_trace(
            go.Scatter(
                  x=df['date']
                , y=df['pageviews']
    #            , fill='tozeroy'
                , name=group_name
            ))

    add_plotly_common(fig)
    add_plotly_title(fig,title)
    show_and_save_plotly_fig(fig,title)
    

def show_and_save_plotly_fig(fig,title):
    fig.show()
    fig.write_html("publish/wd-topic-pageviews/"+topicTitle+"/" + startDay + "_" + endDay + "/plotly_" + title.replace(" ","-") + ".html")