Python Pandas-在多列中添加基于名字和姓氏的新列

Question

Although I'm still a beginner myself, I'm trying to explain some Pandas fundamentals to colleagues who usually manipulate CSV files with Excel.

I hit a wall with my ability to find a "good" answer for solving a given problem I'd like to use as an example.

I have a CSV file like this:

"Id","First","Last"
"109","Karl","Evans"
"113","Louise","Hudson"
"106","Catherine","Johnson"

and I'm importing it into Python like this:

import pandas
df = pandas.read_csv('C:\\example.csv')

I want to add a new column to df called "StartsWithJOrK".

It should say "Yay!" for anyone whose lowercased-first-name OR whose lowercased-last-name starts with a "j" or a "k". It should say "BooHiss" for anyone for whom neither lowercased-name starts with a "j" or a "k".

(It's a rather overwrought example, but I feel like it packs in a lot of things I either don't know how to do or don't know how combine "pythonically.")

What's the most pythonic, fewest-lines-of-code way to do this?

Answer 1

Not the easiest introduction to Pandas...

df['StartsWithJorK'] = 'BooHiss'
starting_letters = ['j', 'k']
df.loc[(df.First.str[0].str.lower().isin(starting_letters)) | 
        df.Last.str[0].str.lower().isin(starting_letters), 'StartsWithJorK'] = 'Yay!'

>>> df
     Id       First     Last StartsWithJorK
0   109        Karl    Evans           Yay!
1   113      Louise   Hudson        BooHiss
2   106   Catherine  Johnson           Yay!

df.First.str[0] finds the first character of the name.

.str.lower() converts this series of letters to lower case.

.isin(starting_letters) checks if each lower case letter is in our list of starting letters, ie 'j' and 'k'.

.loc is for label and boolean based indexing where the column StartsWithJorK is set to Yay! for each matching condition.

Answer 2

If you don't mind importing numpy too, you can do

import numpy as np
import pandas as pd

mask = df['Last'].str.match('[JjKk]') | df['First'].str.match('[JjKk]')
df['StartsWithJOrK'] = np.where(mask, 'Yay!', 'BooHiss')

Output:

    Id      First     Last StartsWithJOrK
0  109       Karl    Evans           Yay!
1  113     Louise   Hudson        BooHiss
2  106  Catherine  Johnson           Yay!

There are other ways of creating the above mask . Here is one:

mask = (df[['First', 'Last']]
            .apply(lambda x: x.str.match('[JjKk]'), axis=1)
            .any(axis=1))

Or, taking a cue from @Alexander's answer's use of .str.lower() :

mask = (df[['First', 'Last']]
            .apply(lambda x: x.str.lower().str.match('[jk]'), axis=1)
            .any(axis=1))