🐼Welcome back to the “Meet Pandas” series (a.k.a. my memorandum for learning Pandas)!🐼

Last time, I discussed grouping and several types of boxplot functions.

Today, I’m going to briefly summarize how to extract a piece of data from a dataframe by specifying some conditions.

Load Example Data

As before, I use the “tips” dataset provided by seaborn. This is a data of food servers’ tips in restaurants with six factors that might influence tips.

import pandas as pd
import seaborn as sns
sns.set()

df = sns.load_dataset('tips')
df

The dataframe should look something like this:

Let’s say we want a subset of records where $2 \leq \mathrm{tip} < 3$ and $\mathrm{day} \in \{\mathrm{Sat},\mathrm{Sun}\}$.

Boolean Indexing

You might solve this problem by Boolean indexing (if you get 28 records, your query is probably correct).

df[ (df['tip']>2) & (df['tip']<3)  & (df['day'].isin(['Sat', 'Sun'])) ]

But, have you felt that this is a little cumbersome and easy to cause mistakes? I have. The common mistakes include:

df[ (2<df['tip']<3)  & (df['day'].isin(['Sat', 'Sun'])) ]
# >> ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

df[ (df['tip']>2) and (df['tip']<3)  and (df['day'].isin(['Sat', 'Sun'])) ]
# >> ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

df[ (df['tip']>2) & (df['tip']<3)  & (df['day'] in ['Sat', 'Sun']) ]
# >> ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Method query

Pandas has a method query, which is easier to read and write than Boolean indexing. This method allows you to specify querying conditions in an SQL-like manner.

df.query('2 < tip < 3 and day in ["Sat", "Sun"]')

I usually prefer query to Boolean indexing because its SQL-like syntax enhances readability and reduces potential errors. But query is not always the better choice. Please note the following:

• If the column name includes spaces or periods, you can’t use query
• pandas.Series objects do not have a method query
• Boolean indexing is a bit faster than query

References

[1] pandas.DataFrame.query — pandas 1.1.1 documentation
[2] pandas.DataFrameの行を条件で抽出するquery | note.nkmk.me
[3] pandasで複数条件のAND, OR, NOTから行を抽出（選択） | note.nkmk.me