CODEWITHSUNDEEP
pythonstringpandasnumpysubstring
nia4life
nia4life

Reputation: 363

Create New Column Based On String

I have a data frame, want to create a column based on the string in column1_sport.

import pandas as pd

df = pd.read_csv('C:/Users/test/dataframe.csv', encoding  = 'iso-8859-1')

Data contains:

column1_sport
baseball
basketball
tennis
boxing
golf

I want to look for certain strings ("ball" or "box") and create a new column based on whether the column contains that word. If the dataframe doesn't contain that word, add "other". See below.

column1_sport    column2_type
baseball         ball
basketball       ball
tennis           other 
boxing           box              
golf             other

Upvotes: 3

Views: 12011

Answers (3)

Ananay Mital
Ananay Mital

Reputation: 1475

df["column2_type"] = df.column1_sport.apply(lambda x: "ball" if "ball" in x else ("box" if "box" in x else "Other"))
df

    column1_sport   column2_type
0        baseball           ball
1      basketball           ball
2          tennis          Other
3          boxing            box
4            golf          Other

Incase you have more complex conditions

def func(a):
    if "ball" in a.lower():
        return "ball"
    elif "box" in a.lower():
        return "box"
    else:
        return "Other"

df["column2_type"] = df.column1_sport.apply(lambda x: func(x))

Upvotes: 2

jpp
jpp

Reputation: 164653

For multiple conditions I suggest np.select. For example:

values = ['ball', 'box']
conditions = list(map(df['column1_sport'].str.contains, values))

df['column2_type'] = np.select(conditions, values, 'other')

print(df)

#   column1_sport column2_type
# 0      baseball         ball
# 1    basketball         ball
# 2        tennis        other
# 3        boxing          box
# 4          golf        other

Upvotes: 5

Vaishali
Vaishali

Reputation: 38415

You can use a nested np.where

cond1 = df.column1_sport.str.contains('ball')
cond2 = df.column1_sport.str.contains('box')
df['column2_type'] = np.where(cond1, 'ball', np.where(cond2, 'box', 'other') )

    column1_sport   column2_type
0   baseball        ball
1   basketball      ball
2   tennis          other
3   boxing          box
4   golf            other

Upvotes: 0

Related Questions