I have a pandas dataframe where each row represents a resume, similar to:

There is an id column, a preprocessed string column with the text of the resume, and 2 columns classifying the applicant's personality (coming from a personality test).

Now, I defined a function that randomly adds an X number of key words to the resume text. However, each color has a different associated list (list containing of typical key words for those personalities) to pull words from:

import random

def addKeyWords(string, color_list):
    resume_word_count = len(string.split())
    percentage = 0.05 # Percentage of total words that needs to be added
    number_of_words_to_be_added = round(int(resume_word_count * percentage))
    list_of_words = random.choices(color_list, k=number_of_words_to_be_added)
    new_string = string   " "   " ".join(list_of_words)
    return new_string 

Now I want to loop through all the rows of the dataframe and apply the function based on the values of color_1 OR color_2.

For example, if either color_1 or color_2 == "orange" then apply the function such as:

df["resume_text_extra"] = df["resume_text"].apply(lambda x: addKeyWords(x, list_orange)) 

However, I can't get it to work with if-else statements within lambda. Any help would be appreciated!

CodePudding user response：

Check Below example code, using np.where. It applies lambda function based on the column value

import pandas as pd

import numpy as np 

df = pd.DataFrame({'col1':[1,2,3,4]})

df['col2'] = np.where(df['col1']<=2, df['col1'].apply(lambda x: x * 2),
              np.where( df['col1'] == 3, df['col1'],
                       df['col1'].apply(lambda x: x * 4)
              ))

df

Output:

CodePudding user response：

How about this?

condition = (df['color_1'] == 'orange') | (df['color_2'] == 'orange')
df[condition] = df[condition].apply(lambda x: addKeyWords(x, list_orange))