How to read file inside class

[buran]

self in the class is the instance itself. If you want to use a instance attribute inside the class you need to use it with self.

class PythonTraining():    
    def read_data(self, filepath):
        base_path = "D:\Data" 
        full_filepath = os.path.join(base_path, filepath, "outsummary2.csv")      
        self.data = pd.read_csv(full_filepath)

    def cal_rowmean(self, v1):         
        self.xmean = self.data['s3'].mean() + v1
        print(self.xmean)

Note, that read_data doesn't need to return self.data.
Also, note that you need to call read_data() before you call cal_rowmean or you will get AttributeError. It may be better if you refactor your class and move what reading data in class __init__ method (you don't have defined one now)

[snippsat]

You lose flexibility and most of the point of this if hardcore most of the stuff inside the class.
It's no so easy with Pandas as often need flexibility to change a lot of stuff(also when read in files).

Here a demo,where take stuff in from outside of the class.
The can continue in normal way to work with df DataFrame object.

Output:
Movie;Year
Seven;1995
The Godfather;1972
Jaws;1975

import pandas as pd
import os

class PythonTraining():
    def __init__(self, base_path, file_name):
        self.base_path = base_path
        self.file_name = file_name

    def read_data(self, sep=';', header='infer'):
        full_filepath = os.path.join(self.base_path, self.file_name)
        self.data = pd.read_csv(full_filepath, sep=sep, header=header)


if __name__ == '__main__':
    # Now from outside give path and filename
    base_path = r'E:\div_code\home'
    file_name = 'movies.csv'
    df = PythonTraining(base_path, file_name)

    # Give paramater to csv read
    sep = ';'
    #header = None
    df.read_data(sep)
    df = df.data

As you use Spyder here a Screenshot,see that i can continue to work with df object in IPython console.