This post is the second part of filtering a pandas DataFrame. The first part is in the post called Filtering a pandas DataFrame. In this second part I look at multiple conditions.
Let’s use a simple example by creating a pandas DataFrame manually.
2 | data1 = {'firstname': ['Bob', 'Sally', 'Suzie', 'Rohan'], |
3 | 'amount': [12, 67, 33, 41]} |
4 | df1 = pd.DataFrame(data1) |
1 | df2 = df.loc[df["amount"] >= 35] |
The above query returns the Sally and Rohan rows.
2 | df2 = df.loc[df["firstname"].str.contains("S")] |
The above query returns Sally and Suzie. The contains function is case sensitive. If I had uses ‘s’ instead, it would not have returned any rows.
Using query method
3 | df3 = df.query("firstname.str.contains('a', case=False)") |
The above query returns the Sally and Rohan rows.
2 | df4 = df.query("firstname.str.contains('a', case=False) or firstname.str.contains('z')") |
The above returns the Sally, Suzie and Rohan rows, which is what we expected.
2 | df5 = df.query("firstname.str.contains('a', case=False) or amount == 12") |
The above query returns the Bob, Sally and Rohan rows, which is what we expected. You’ll notice that the quotes need to be different.
2 | df6 = df.query("firstname.str.contains('a', case=False) and amount == 67") |
The above query returns the Sally row.
