I have the following dataframe, based on data i pulled from my database:
| date | event_type | count |
|---|---|---|
| 2022-05-10 | page_view | 3 |
| 2022-05-11 | cart_add | 2 |
| 2022-05-11 | page_view | 2 |
| 2022-05-12 | cart_add | 1 |
| 2022-05-12 | cart_remove | 1 |
| 2022-05-12 | page_view | 2 |
| 2022-05-13 | cart_remove | 2 |
| 2022-05-13 | page_view | 1 |
| 2022-05-14 | cart_add | 2 |
| 2022-05-14 | page_view | 5 |
Basically I am tracking 3 things on my website:
- when a user views a product page
- when a user adds a product to their cart
- when a user removes a product from their cart
I’m tracking how often each of these events happens in a day and I want to then graph them all on a single line chart. In order to do that, I think I need to make it look something more like this:
| date | page_views | cart_adds | cart_removes |
|---|---|---|---|
| 2022-05-10 | 3 | 0 | 0 |
| 2022-05-11 | 2 | 2 | 0 |
| 2022-05-12 | 2 | 1 | 1 |
| 2022-05-13 | 1 | 0 | 2 |
| 2022-05-14 | 5 | 2 | 0 |
I am very new to pandas and not even sure if this library is what I should be using. So forgive my cluelessness, but how do I make dataframe1 look like dataframe2?
>Solution :
df.pivot(columns='event_type', index='date').fillna(0)
Output:
count
event_type cart_add cart_remove page_view
date
2022-05-10 0.0 0.0 3.0
2022-05-11 2.0 0.0 2.0
2022-05-12 1.0 1.0 2.0
2022-05-13 0.0 2.0 1.0
2022-05-14 2.0 0.0 5.0