How to get the rank of one numpy array based on intervals from another array?-CodePudding

I have two numpy arrays:

arr = np.array([10,80,10,20,60,50,80,100])
intervals = np.array([20,35,60,100]) # this is always increasing


# we can get ranks from intervals
ranks = np.arange(len(intervals))   1 # array([1, 2, 3, 4])

Here, if the values is less than 20, it has rank 1, if 20<x<=35 rank2 and so on.

How to get the following answer? Either the dictionary or the array is fine.

required_answer_dict = {10:1, # <= 20
            80:4, # <=100
            10:1, # <= 20
            20:1, # <=20
            60:3, # <=60
            50:3, # <=60
            80:4, # <=100
            100:4 # <=100
           }

required_answer = [1,4,1,1,3,3,4,4]

CodePudding user response：

You can use np.digitize for this:

binned = np.digitize(arr - 1, intervals)   1
ans_dict = dict(zip(arr, binned))

Output:

>>> ans_dict
{
    10: 1,
    80: 4,
    20: 1,
    60: 3,
    50: 3,
    100: 4
}

CodePudding user response：

Reshape into a long format and compare each element to the intervals. The argmax() will return the values you are looking for.

import numpy as np
arr = np.array([10,80,10,20,60,50,80,100])
intervals = np.array([20,35,60,100]) # this is always increasing

(arr.reshape(-1,1) < intervals).argmax(axis=1) 1

Output

array([1, 4, 1, 2, 4, 3, 4, 1], dtype=int64)