calculationg the mean from each Dataframe column-CodePudding

I've to write a function (column_means), that calculates the mean of each column from Dataframe and give me a list of means at the end. I'm not allowed to use the mean function .mean(), so I'm implementing the general formula of the mean: sum(x_i)/Number of elements.

This is my code:

df = pd.DataFrame({'a':[1,2,3], 'b': [4,5,6]})
def column_means(df):
    means  = [] 
    for i,n in zip(df.columns,  df.shape[0]):
        means [n] = sum(df[i])/ df.shape[0]
    return means

It doesn't work as intended. could you please help me and tell me, what are my mistakes?

Thank you in advance.

CodePudding user response：

You are iterating over int in zip function, as df.shape[0] is returning single integer and not an iterable datatype.

So you can simply do as following:

def column_means(df):
    means = []
    for i in df.columns:
        means.append(sum(df[i]) / df.shape[0])
    return means

And if you want mean to be just an integer instead of float, you can just do sum(df[i]) // df.shape[0]

I hope this answers your question.

CodePudding user response：

Do you want the mean of each column? You have to be careful if they don't have the exact same length:

import pandas as pd

df = pd.DataFrame({'a':[1,2,3], 'b': [4,5,6]})

def column_means(df):
    means  = [] 
    for i,n in enumerate(df.columns):
        means.append(sum(df[n])/len(df[n]))
    return means

print(column_means(df))

You can also use the mean method of pd DataFrame

df.mean()

CodePudding user response：

change the first df.shape[0] to df.indexand the assignment line.

def column_means(df):
   means  = [] 
   for i,n in zip(df.columns,  df.index):
      means.append(sum(df[i])/ df.shape[0])
   return means

CodePudding user response：

If the only thing you're not allowed to use is the df.mean() function, then you could do:

def column_means(df):
    return df.sum(axis=0).div(df.shape[0]).to_list()

Sum over the columns, divide the result by the number of rows, and convert it to a list.