How To Quantitatively Measure Goodness Of Fit In Scipy?

I am tying to find out the best fit for data given. What I did is I loop through various values of n and calculate the residual at each p using the formula ((y_fit - y_actual) / y_

Solution 1:

Probably the most commonly used goodness-of-fit measure is the coefficient of determination (aka the R value).

The formula is:

where:

Here, y_i refers to your input y-values, f_i refers to your fitted y-values, and ̅y refers to the mean input y-value.

It's very easy to compute:

# residual sum of squaresss_res = np.sum((y - y_fit) ** 2)

# total sum of squaresss_tot = np.sum((y - np.mean(y)) ** 2)

# r-squaredr2 = 1 - (ss_res / ss_tot)