import json
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import matplotlib.ticker as mticker
from datetime import datetime, timezone

import seaborn as sns
sns.set_theme()
sns.reset_defaults()

plt.rcParams['figure.figsize'] = (10, 5)

df = pd.read_json('netblocks-entity.json', lines=True, )
df = df[['url', 'date', 'content']]

df = df.set_index('date')

df.to_csv('netblocks.csv')
df[['url']].to_csv('netblocks-simple.csv')

def plot_freq(freq, name, n):

    resampled = pd.DataFrame({'count': df['url'].resample(freq).count().sort_index()})
    resampled['date'] = resampled.index.date

    plt.figure()
    fig, ax = plt.subplots()

    resampled.plot(x='date', y='count', kind='bar', ax=ax).legend([f'{name.title()} internet disruptions'])
    #ax.plot(resampled.date, resampled.count, kind='bar')
    #ax.axhline(y=3, color='green', linestyle='--', label='count students at initial start')

    # Fix date x tick labels
    # Fill all tick labels with empty values:
    ticklabels = [''] * len(resampled)
    # Divide series into n parts and show only label for start of part:
    # n = 18
    skip = len(resampled) // n
    # Format *only* every nth tick label as %Y-%m-%d
    ticklabels[::skip] = resampled['date'].iloc[::skip] # .strftime('%Y-%m-%d')
    ax.xaxis.set_major_formatter(mticker.FixedFormatter(ticklabels))
    # Auto-rotate labels 45deg, from https://matplotlib.org/2.0.2/users/recipes.html
    fig.autofmt_xdate()

    # Unused
    #fig.autofmt_xdate()
    #ax.fmt_xdata = mdates.DateFormatter('%Y-%m-%d')

    # Vertical line
    date = datetime(2022, 4, 10, 0, 0, tzinfo=timezone.utc)
    pos = resampled.index.get_loc(
        resampled.index[resampled.index == date][0]
    )
    #ax.axvline(pos, color='r', lw=1, label='Important date')

    # Doesn't do much?
    #ax.xaxis_date()

    ax.set_xlabel('')
    #plt.legend()

    filename = f'netblocks-{name}-disruptions'
    plt.savefig(f'{filename}.svg')
    plt.savefig(f'{filename}.png')
    plt.show()

plot_freq('1d', 'daily', 20)
plot_freq('1w', 'weekly', 15)

<Figure size 1000x500 with 0 Axes>

<Figure size 1000x500 with 0 Axes>

import plotly.express as px

fig = px.bar(df['url'].resample('1w').count(), labels={'value': 'Disruptions'}).update_layout(
    autosize=True,
    margin=dict(l=0, r=0, b=0, t=0, pad=0),
).update_layout(showlegend=False)
fig.write_image("netblocks-weekly-disruptions.svg")
fig.show()

df_plotly = df['url'].resample('1d').count().rolling(7).mean()

fig = px.bar(
    df_plotly,
    #title='Weekly disruptions',
    text=df_plotly,
    labels={'date': '', 'value': 'Disruptions'},
)
fig.update_traces(textposition='outside')
# Hide axis labels
fig.update_layout(yaxis_visible=True, yaxis_showticklabels=True,
                  xaxis_visible=True, xaxis_showticklabels=True)
# Hide legend
fig.update_layout(showlegend=False)

# Disable margins and padding
fig.update_layout(
    autosize=True,
    margin=dict(l=0, r=0, b=0, t=0, pad=0),
)

fig.write_image("netblocks-disruptions-7day-rolling-mean.svg")

fig.show()