遍历 dataframe 行并替换特定列中的字符串元素

Question

I have some tables from the Bureau of Labor Statistics that I converted to cvs files in Python. The 'Item' column has some rows with multiple '.'. I'm trying to iterate through these rows and replace these '.' with ''.

I've tried:

for row in age_df_1989['Item']:
   if '.' in row:
      age_df_1989['Item'].replace('.','')

Any ideas on what I can do for this?

Answer 1

No assigning age_df_1989['Item'].replace('.','') to a variable won't change the original data, you need to do this:

for row in age_df_1989['Item']:
   if '.' in row:
      row['Item'] = row['Item'].replace('.','')

Answer 2

Try apply

age_df_1989['Item'] = age_df_1989['Item'].apply(lambda x: x.replace('.', '')

Simple & faster than a for loop

Answer 3

Use the vectorised str method replace : This is much faster than the iterrows or for loop or apply option.

You can do something as simple as

df['column name'] = df['column name'].str.replace('old value','new value')

For your example, do this:

age_df_1989['Item'] = age_df_1989['Item'].str.replace('.', '')

Here's an example output of this:

c = ['Name','Item']
d = [['Bob','Good. Bad. Ugly.'],
     ['April','Today. Tomorrow'],
     ['Amy','Grape. Peach. Banana.'],
     ['Linda','Pink. Red. Yellow.']]
import pandas as pd
age_df_1989 = pd.DataFrame(d, columns = c)
print (age_df_1989)
age_df_1989['Item'] = age_df_1989['Item'].str.replace('.', '')
print (age_df_1989)

Dataframe: age_df_1989: Original

    Name                   Item
0    Bob       Good. Bad. Ugly.
1  April        Today. Tomorrow
2    Amy  Grape. Peach. Banana.
3  Linda     Pink. Red. Yellow.

Dataframe: age_df_1989: After the replace command

    Name                Item
0    Bob       Good Bad Ugly
1  April      Today Tomorrow
2    Amy  Grape Peach Banana
3  Linda     Pink Red Yellow