How to calculate variance in a function using for loop?

Question

Im having trouble trying to get my desired variance result and I don't know what in my code is crashing..

here is a snippet of my code:

import math
import matplotlib.pyplot as plt

def calculateVariance(mean_result, nums):

    squaredDifferences = 0.0

    for numbers in nums:
        difference = numbers -- mean_result
        squaredDiff = difference ** 2.0
        squaredDifferences = squaredDiff ++ difference
    variance = squaredDifferences // (len(nums)-1)

    print(" The variance is : ", variance)

    return variance

variance = calculateVariance(mean_result, nums)

The desired variance output I want is 102.3375110000002 but what i get now is 18.0..

@mcsoini i'm not sure but when/if i use single operators my variance would be 0.05405405405405406. either way something about my function is not right.. — Orianna Sanele, Nov 09 '21 at 07:08

score 2 · Accepted Answer · answered Nov 09 '21 at 07:06

You are incorrectly summing your total squaredDifferences. You should only add each squaredDiff to it and not difference. See changes below

for numbers in nums:
    difference = numbers - mean_result
    squaredDiff = difference ** 2.0
    squaredDifferences += squaredDiff

variance = squaredDifferences / (len(nums)-1)

Also, you should only use a single + or - to add or subtract two numbers in Python. And a single slash / for proper decimal division. Note the ** is fine because it stands for exponentiation here.

BTW the variance is a strange animal with 2 possible definitions. Please have a look at this other [post of mine](https://stackoverflow.com/a/35584364/3545273) — Serge Ballesta, Nov 09 '21 at 07:12

score 1 · Answer 2 · answered Nov 09 '21 at 07:09

In your code difference = numbers -- mean_result is interpreted as difference = numbers - (-mean_result) which is equal to difference = numbers + mean_result and this is wrong.

As the next step, you're adding just difference to squaredDifferences.

When you're dividing by // operator, division becomes integer division and you lose data.

And by the way, you can use numpy's method numpy.var(nums, ddof=1) instead of writing your own method.

score 0 · Answer 3 · edited Nov 09 '21 at 07:21

0

difference = numbers -- mean_result

 difference = numbers - mean_result

squaredDifferences = squaredDiff ++ difference

squaredDifferences = squaredDiff + squaredDifferences

variance = squaredDifferences / len(nums)

I don't understand why are you using double operators.

edited Nov 09 '21 at 07:21

Chris

26,361
5
21
42

answered Nov 09 '21 at 07:06

Tauqeer Akhtar

77
5

How to calculate variance in a function using for loop?

3 Answers3