A Python program that aggregates time usage from icalendar data

Package installation

First, pip install icalendar is.

Next, create the items to be used for analysis in Excel.

Enter the item of the unit you want to analyze on the first line, and the sub-items of each analysis unit on the second and subsequent lines. This sub-item becomes the search keyword for iCalendar data.

I will also name it item table .xlsx and save it.

Next, prepare the icalendar data to be analyzed.

I think there are various ways to prepare. Exporting from Outlook is easy, isn't it?

Next is the source.

To be honest, I don't think it's elegant at all, but since I'm a beginner, please forgive me, or please give me some guidance.

-First read the item table and create a dataframe for data aggregation of the same size -One after another, icalendar data is read, the corresponding coordinates are found from the item table.xlsx, and the calculated time is added to the corresponding data frame for aggregation.

Waste of memory? I think there are many things like double loops are not elegant, so please comment.

import numpy as np
import pandas as pd
import icalendar as ic
from datetime import timedelta


def calculate_time(event):
    start = event['DTSTART'].dt
    end = event['DTEND'].dt

    time = end -start

    return time.total_seconds() / 3600


def main():
    labels = pd.read_excel('./Item list.xlsx')
    print(labels)

    file = open('./testdata.ics', 'r', encoding='utf-8')
    cal = ic.Calendar.from_ical(file.read())

    data = pd.DataFrame(np.zeros((labels.shape[0], labels.shape[1])))
    data.columns = label.columns

    for event in cal.walk('vevent'):
        time = calculate_time(event)
        summary = str(event.get('summary'))

        for i in range(0, labels.shape[0]):
            for j in range(0, labels.shape[1]):
                if str(labels.iloc[i, j]) in summary:
                    data.iloc[i, j] = data.iloc[i, j] + time


    print(data)

if __name__ == '__main__':
    main()

If you don't know what you are doing, please ask. I haven't graphed it yet, but at this point it should be easy to graph.

An example of graphing

I'll make it a pie chart.

data.sum().plot.pie(legend=False, autopct="%.1f%%", textprops={'fontproperties': fp, 'fontsize':20})
plt.axis('equal')

I try to display the percentage as autopct = "% .1f %%" so that legend = False does not cause annoying legends.

The point when using Japanese fonts was that it took time to find out how to write textprops = {'fontproperties': fp,'fontsize': 20}). I couldn't find any information on how to specify a Japanese font when using a pie chart with a python plot, and it took a long time.

plt.axis ('equal') prevents the pie chart from collapsing.

References

・ Https://gist.github.com/lukasmartinelli/9021795