CODEWITHSUNDEEP
pythonpandascsv
sshifeed
sshifeed

Reputation: 11

how to use pandas and get csvfile value

I have a csv file. test.csv

1,1,1,1

I want to load this as variables.

My code:

import pandas as pd
ch_file=pd.read_csv("test.csv")
ch1_value=ch_file.columns[0]
ch2_value=ch_file.columns[1]
ch3_value=ch_file.columns[2]
ch4_value=ch_file.columns[3]

print(ch1_value, ch2_value, ch3_value, 
 ch4_value)

Output is 1,1.1,1,1.1

Why does it not print 1,1,1,1?
Where is the .1 coming from?

Upvotes: 1

Views: 37

Answers (2)

Corralien
Corralien

Reputation: 120391

You have 2 problems here:

  • When you use read_csv the first line is interpreted as the column headers. You have to pass header=None if you have not.

  • ch_file.columns[2] return the column name and not the value(s). Instead, use .iloc.

ch_file = pd.read_csv('test.csv', sep=',', header=None)
ch1_value=ch_file.iloc[0, 0]
ch2_value=ch_file.iloc[0, 1]
ch3_value=ch_file.iloc[0, 2]
ch4_value=ch_file.iloc[0, 3]
print(ch1_value, ch2_value, ch3_value, ch4_value)

Output:

1 1 1 1

Upvotes: 1

mozway
mozway

Reputation: 260300

What do you want to extract? Just the numbers?

pandas.read_csv rewrites the column names to be unique, so it will add .1, .2, etc. for this. In theory you could pass the mangle_dupe_cols=False option, but it is not supported yet.

An alternative it to pass header=None and to get the values of the first row:

pd.read_csv('test.csv', header=None).iloc[0]

In your case:

ch1_value, ch2_value, ch3_value, ch4_value = pd.read_csv('test.csv', header=None).iloc[0]

But, honestly, do you really need pandas for this? You could do it directly in pure python:

with open('test.csv') as f:
    ch1_value, ch2_value, ch3_value, ch4_value = map(int, f.readline().split(','))

Upvotes: 1

Related Questions