[FIXED] Select rows from pandas that contain given string in list

Issue

I want to select the rows from dataframe that have "apple" in their column, but the "PHRASE" column should not be considered as string but as a list with "|" separator. Phrase column can have None/NaN values.

import pandas as pd

df = pd.DataFrame({"PHRASE": ["apple and bannaa|bannana", None, "apple", "orange|bannana", "orange|apple|bannana"]})

print(df)

0  apple and bannaa|bannana
1                     apple
2            orange|bannana
3      orange|apple|bannana

I want to select rows that have apple in them, but the row should be considered as list with "|" as separator.

Desired output:

1                     apple
3      orange|apple|bannana

I tried to do this

result = df[df.PHRASE.str.split("|").str.contains("apple")]

but it does not work.

Solution

You can still use str.contains but with a regex pattern

df[df['PHRASE'].str.contains(r'\bapple(?:\||$)', na=False)]

---

                 PHRASE
1                 apple
3  orange|apple|bannana

Answered By - Shubham Sharma