CODEWITHSUNDEEP
pythonnumpy
bananagator
bananagator

Reputation: 581

Increase size of an array by filling the values inbetween with the previous value

I have a array of size 19. I want to increase its size to 30(new_array).

size =len(VM)    
index = np.linspace(1,size,size)/size
index30 = np.linspace(1,30,30)/30
new_array = np.empty(shape=30)

VM is a array of size 19

index

[0.05263158 0.10526316 0.15789474 0.21052632 0.26315789 0.31578947
 0.36842105 0.42105263 0.47368421 0.52631579 0.57894737 0.63157895
 0.68421053 0.73684211 0.78947368 0.84210526 0.89473684 0.94736842
 1.        ]

index30

[0.03333333 0.06666667 0.1        0.13333333 0.16666667 0.2
 0.23333333 0.26666667 0.3        0.33333333 0.36666667 0.4
 0.43333333 0.46666667 0.5        0.53333333 0.56666667 0.6
 0.63333333 0.66666667 0.7        0.73333333 0.76666667 0.8
 0.83333333 0.86666667 0.9        0.93333333 0.96666667 1.        ]

To fill the new_array:

If we consider first two elements of index array 0.05263158 0.10526316, all the values of index30 array between these two values, their corresponding locations in the new_array should be filled with the value of VM array that corresponds to location of value 0.05263158 in index array and so on. I can use for loops to do this but i am looking for more efficient way to do this?.

Input:

[ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]

expected output:

[nan,1,1,2,3,3,4,5,5,6,6,7,8,8,9,10,10,11,12,12,13,13,14,15,15,16,17,17,18,19]

Upvotes: 1

Views: 347

Answers (2)

Halbeard
Halbeard

Reputation: 121

You can use a boolean mask to determine where the value from the old array can be inserted into the new one. This is faster than a for loop by creating a correlation matrix with np.newaxis in the index and then np.argwhere(mask) to get the mapping from the old array to the new.

import numpy as np

VM = np.arange(1, 20)
size =len(VM)
index = np.linspace(1,size,size)/size
index30 = np.linspace(1,30,30)/30
new_array = np.empty(30)

mask = index30[:, np.newaxis] >= index[np.newaxis, :]
mask[:,:-1] = np.logical_and(
        mask[:,:-1],
        index30[:, np.newaxis] < index[np.newaxis, 1:]
    )

index_map = np.argwhere(mask)
new_array[index_map[:,0]] = VM[index_map[:,1]]
print(new_array)

The output of this code is

[4607182418800017408  1  1  2  3  3  4  5  5  6  6  7  8  8  9 10 10 11 12 12 13 13 14 15 15 16 17 17 18 19]

where the first number is the uninitialized value from the empty array.

Note that you cannot get the index map completely by simply doing

index_map = np.argwhere(np.logical_and(
    index[np.newaxis, :-1] <= index30[:, np.newaxis],
    index[np.newaxis, 1:] > index30[:, np.newaxis]
))

as it will omit the last entry. Since two arguments in np.logical_and have to have the same dimensions we need the second statement for manipulating the mask.

Upvotes: 1

Ben.T
Ben.T

Reputation: 29635

EDIT: You can use np.piecewise. With np.less_equal and np.outer to create the condlist and np.append a nan to VM to create the values of the funclist such as:

new_array = np.piecewise( x = index30, 
                          condlist = np.less_equal.outer(indexVM, index30), 
                          funclist = np.append(VM,np.nan))

If you are willing to use pandas, you do a reindex with the method 'ffill'.

import numpy as np
import pandas as pd

VM = np.arange(1,20)
size = len(VM)    
indexVM = np.linspace(1,size,size)/size
index30 = np.linspace(1,30,30)/30 

new_array = pd.Series(VM, index=indexVM).reindex(index30, method='ffill').values
print (new_array)
array([nan,  1.,  1.,  2.,  3.,  3.,  4.,  5.,  5.,  6.,  6.,  7.,  8.,
        8.,  9., 10., 10., 11., 12., 12., 13., 13., 14., 15., 15., 16.,
       17., 17., 18., 19.])

Upvotes: 3

Related Questions