如何用pandas数据框打开csv文件

Question

There is a CSV format file with three column dataframe. The third column has long text. This error message occurred, when i tried to open the file using pandas.read_csv

message : UnicodeDecodeError: 'utf-8' codec can't decode byte 0xa1 in position 0: invalid start byte.

But there is no problem opening the file with

with open('file.csv', 'r', encoding='utf-8', errors = "ignore") as csvfile:

I don't know how converting this data to dataframe and i don't think pandas.read_csv handle this error properly.

So, how can i open this file and get dataframe?

Answer 1

Try this:

Open the cvs file in a text editor and make sure to save it in utf-8 format.

Then read the file as normal:

import pandas
csvfile = pandas.read_csv('file.csv', encoding='utf-8')

Answer 2

I would try using the built-in csv reader then put the data into pandas.

import csv
with open('eggs.csv', newline='') as csvfile:
     spamreader = csv.reader(csvfile, delimiter=' ', quotechar='|')
     for row in spamreader:
         print(', '.join(row))

If this doesn't work, then at least you can confirm that it is a csv issue and not a pandas issue choking on encodings.

The other recommendation is to ensure you are using Python 3.x that handles encoding issues much better than 2.7.

If you can provide your sample, I can test it myself and update my answer accordingly.

Answer 3

You can try another option for encoding as "ISO-8859-1"

In your case:

with open('file.csv', 'r', encoding = 'ISO-8859-1', errors = "ignore") as csvfile:

or try this:

import pandas as pd
data_file = pd.read_csv("file.csv", encoding = "ISO-8859-1")
print(data_file)