[FIXED] Skew array toward maximum value

Issue

Say I have an array

A = np.arange(1,11)
A = np.array([1,2,3,4,5,6,7,8,9,10])

and I want to apply a skew to it to shift values towards the maximum, whilst keeping the upper and lower limits the same, to achieve something like this:

B = myfunc(A)
B = np.array([1,5,7.5,8.7,9.2,9.5,9.75,9.85,9.95,10])

How would I best go about this?

Note, the values chosen are random values from my head to give an idea of the general idea :)

Thanks

Solution

You can write a function which skews the values using a logarithmic or exponential function. In your case, you want a transformation that gets closer to the maximum value while preserving the original scale.

Here's an approach using a logarithmic function to achieve a skewed transformation:

import numpy as np

def myfunc(arr):
    min_val = arr.min()
    max_val = arr.max()
    range_val = max_val - min_val
    
    # Apply a logarithmic function to skew the values
    log_base = 10  # Adjust this base for desired skew
    shifted = arr - min_val + 1  # Shift values to avoid log(0)
    log_values = np.log(shifted) / np.log(log_base)
    
    # Scale the log-transformed values back to the original range
    result = (log_values / log_values.max()) * range_val + min_val
    
    return result

A = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

B = myfunc(A)
print(B)

This function myfunc uses a logarithmic transformation to skew the values towards the maximum while preserving the upper and lower limits. The log_base variable is used to control the extent of the skew.

This code shifts the values to avoid taking the logarithm of zero, applies the logarithmic transformation, normalizes the values between 0 and 1, and scales them back to the original range.

You can adjust the log_base value to achieve different degrees of skew towards the maximum while ensuring the upper and lower limits remain the same.

Answered By - nithinks