将单引号替换为双引号 python pandas dataframe

Question

I want to replace single quote(') to double quote(") to make it proper json column value in python dataframe.

eg csv file looks like...

Unit Id Batch Id                               Items prod
A108    qa120  {'A': 123, 'B': 342, 'C': 454}   
P258    re015  {'A': 124, 'B': 234, 'C': 343} 

I'm reading these values from csv to pandas dataframe. I tried several ways, but no luck.

df.replace("'",'"',inplace=True)
df.['<column_name>'].str.replace(r"[\',]",'"')
df = df['<column_name>'].str.replace(r"[\',]",'"')

Thanks for your help in advance.

Answer 1

If the problem is converting the single quote to double quotes without the restraint of doing it after you read it into a dataframe - you could change the .csv file before you read it into a dataframe:

$ sed -i "s/'/\\"/g" file_name.csv

If you have to replace them after you read them into a dataframe, try the solution mentioned in this post :

df.replace({'\\'': '"'}, regex=True)

Answer 2

Use str.replace .

If you want to update a column on a DataFrame, such as this

And let's say that you want to remove the double quotes from the first column.

Just do the following

df[0] = df[0].str.replace(r"[\"]", r"'")

Here is the final result

Answer 3

You can convert values to dictionaries like:

import ast

df['<column_name>'] = df['<column_name>'].apply(ast.literal_eval)

But if input data are json file (string), better is use json_normalize .

Answer 4

Looks like you need.

import pandas as pd
import json
import ast


df = pd.DataFrame({"Unit Id": ["A108", "P258"], "Batch Id": ["qa120", "re015"], "Items prod": ["{'A': 123, 'B': 342, 'C': 454}", "{'A': 124, 'B': 234, 'C': 343}"]})
df["NEW"] = df["Items prod"].apply(ast.literal_eval).apply(json.dumps)
print(df)

Output:

  Batch Id                      Items prod Unit Id  \
0    qa120  {'A': 123, 'B': 342, 'C': 454}    A108   
1    re015  {'A': 124, 'B': 234, 'C': 343}    P258   

                              NEW  
0  {"A": 123, "C": 454, "B": 342}  
1  {"A": 124, "C": 343, "B": 234}