CODEWITHSUNDEEP
pythonpandas
NineWasps
NineWasps

Reputation: 2253

Pandas: parse columns from csv

I have data in csv file without header. I need to parse some columns.

A part of data:

-1.0,-0.0246259814315,1174.60023796
 1.0,-0.978057706084,1083.19880269
-1.0,0.314271994507,-1472.97760911
-1.0,0.179751565771,231.017267343
1.0,-1.26254374278,-778.271726463
-1.0,0.249969939456,-52.8014826538
1.0,-1.87039747875,-324.235348241

I need to load only second and third columns. I use train_X = pd.read_csv("perceptron-train.csv", sep=',', parse_dates=[1], usecols=[2, 3]) but it returns IndexError: list index out of range

Upvotes: 0

Views: 1179

Answers (1)

EdChum
EdChum

Reputation: 393933

IIUC indices are zero-based so you need:

train_X = pd.read_csv("perceptron-train.csv", sep=',', parse_dates=[1], usecols=[1, 2])

Also I don't know if this also means you need to change your date col:

train_X = pd.read_csv("perceptron-train.csv", sep=',', parse_dates=[0], usecols=[1, 2])

However, looking at your data I don't understand how to interpret the first or second column as a datetime as they look weird

Upvotes: 1

Related Questions