Stacked bar plotting dataframe groups

Question

I'm trying to plot stacked bar chart from a dataframe for several hours. I'm sorry if this is a bare question, but I just can't make it work, I need help.

My dataframe looks like this:

                                 _id        date                              news_source
0   2715eeada6726024df20e6938ef09f64  2019-12-23                    airport-suppliers.com
1   d068a3d0b24d2a348ff8c8a856aba86c  2019-12-23                    airport-suppliers.com
17  552d7bb9f7d3fd689dd308dc7650baac  2019-12-23                    airport-suppliers.com
20  82be33a041204fd008ba5093607310f6  2019-12-23                    airport-suppliers.com
21  4044907f5b6d5610ec59a03c75e0554c  2019-12-23  airportsinternational.keypublishing.com
22  db4e1e4d1246abc3304e5d77688424dc  2019-12-23  airportsinternational.keypublishing.com
23  b7f57b63218190d249d19624bbdcb520  2019-12-23           internationalairportreview.com
27  84d5377bd8755a685100e408140c4ab1  2019-12-23           internationalairportreview.com
28  8289a1c1b3fa3f618c332d61023eae00  2019-12-16               passengerterminaltoday.com
29  f4f020f09ee5f95499a26c43cfd82d2d  2019-12-16  airportsinternational.keypublishing.com
..                               ...         ...                                      ...
59  a18388a1c77889bdbe6aaa9238a8d21a  2019-12-16                    airport-suppliers.com
62  5cd894a9fa587ab4267adfd23f01e1c4  2019-12-16  airportsinternational.keypublishing.com
66  bb7d05d61f999b1f0b317d21c6c23c0c  2019-12-16  airportsinternational.keypublishing.com
70  f49b9ce330198aec666cb90275d293b2  2019-12-16           internationalairportreview.com
71  af893db09fad9335413ce5c325ced712  2019-12-16               passengerterminaltoday.com
72  e21dc60cfda457b03a6dba6ab44aa3b1  2019-12-16               passengerterminaltoday.com
81  963760af4b4653d175902f4d6285ff0a  2019-12-16               passengerterminaltoday.com
82  778b572be28fd25f394cfa41bbc5aa4a  2019-12-16                    airport-suppliers.com

The final plot I want to show is like this, but instead of strategies there will be weekly dates, news_source instead of Products, and counts is the same.

What I tried is groupby by date and news_source, then counting them. Then the rest of my work just got messed up and in the end I couldn't make it to get in a format like the example in this. Also, the amount of unique news_source, date may change over time, so I'm avoiding hardcoding things as much as I can.

The grouping:

groups = df.groupby(['date', 'news_source'])["_id"].count()

If you need them as dictionary:

counts = defaultdict(dict)
for index, count in zip(groups.index, groups):
    try:
        counts[index[0]][index[1]] += count
    except KeyError:
        counts[index[0]][index[1]] = count

Output is:

{'2019-12-16': {'airport-suppliers.com': 9,
                'airportsinternational.keypublishing.com': 12,
                'internationalairportreview.com': 19,
                'passengerterminaltoday.com': 21},
 '2019-12-23': {'airport-suppliers.com': 21,
                'airportsinternational.keypublishing.com': 2,
                'internationalairportreview.com': 5}}

If you know how to do it properly, any help will be appreciated, thanks.

Here is the code to generate minimal reproducible example:

import pandas as pd

dates = ['2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-23', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16', '2019-12-16']

sources = ['airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'passengerterminaltoday.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'airport-suppliers.com', 'passengerterminaltoday.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'airport-suppliers.com', 'passengerterminaltoday.com', 'airport-suppliers.com', 'airport-suppliers.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airport-suppliers.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'airportsinternational.keypublishing.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'passengerterminaltoday.com', 'airport-suppliers.com', 'airport-suppliers.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com', 'internationalairportreview.com']

df = pd.DataFrame({"date": dates, "news_source": sources})  

Answer 1

How about this? I added counts for your data:

df1 = df.groupby(['date', 'news_source']).size().reset_index().rename(columns={0:'count'})

Then, I used pd.crosstab, set the following index, columns, and values parameters. Then include an aggfunc, which is sum() in this case.

pd.crosstab(index=df1['date'], columns=df1['news_source'], values=df1['count'], aggfunc=sum).plot.bar(stacked=True)

Result: