plot_message_times.py

import csv
import matplotlib.pyplot as plt
import pandas as pd
from tqdm import tqdm
import glob
from dateutil.parser import parse

# Read messages.csv in all data/messages/* directories
message_files = glob.glob('data/messages/*/messages.csv')

# Create a list of all message times
message_times = []
for file in tqdm(message_files):
	with open(file, 'r') as f:
		reader = csv.reader(f)
		next(reader)
		for row in reader:
			message_times.append(parse(row[1]))

# sort the message times
message_times.sort()

print(message_times[0:10])

# Create a list of times of day
times = [t.hour + t.minute/60 for t in message_times]

# Plot the times
plt.scatter(times, message_times, alpha=0.5, s=0.5)
plt.xlabel('Time of day')
plt.ylabel('Date')

# scale the x axis to be 24 hours
plt.xlim(0, 24)

# scale the y axis to be the first and last date
plt.ylim(message_times[0], message_times[-1])

# set the layout to be tight
plt.tight_layout()

# show months on the y axis
plt.yticks(pd.date_range(message_times[0], message_times[-1], freq='ME'))

# show each hour on the x axis
plt.xticks(range(24))

# invert the y axis so that the first date is at the top
plt.gca().invert_yaxis()

# set layout to be tall
plt.gcf().set_size_inches(10, 20)

# increase resolution
plt.gcf().set_dpi(300)

# save the plot to a file
plt.savefig('message_times.png', bbox_inches='tight', dpi=300)