CODEWITHSUNDEEP
pythonarrayspython-3.xnumpycombinations
Eli
Eli

Reputation: 4936

Numpy combine two MaskedArrays with OR operation

I have two MaskedArray objects, with the same length:

>> grades_passed
   [90 -- 88 62 -- 100]
>> grades_not_passed
  [-- 54 -- -- 34 --]

And I'm trying to combine them to get:

[90 54 88 62 34 100]

I tried some manipulations like:

total_final = grades_passed + grades_not_passed

or

total_final = ma.mask_or(grades_passed, grades_not_passed)
total_final = ma.concatenate(grades_passed, grades_not_passed)

But none of them do the trick

Upvotes: 1

Views: 123

Answers (2)

Felix
Felix

Reputation: 1905

The following works, as long as there is no overlap of the two masks:

grades = grades_passed.filled(1) * grades_not_passed.filled(1)

The parameter 1 temporarily sets the fill_value of the two variables, so masked values take on 1 as value to be filled.

Upvotes: 2

Mark Moretto
Mark Moretto

Reputation: 2348

Convert the arrays to lists and create dictionaries with index values (this assumes that the two arrays won't have a value in the same index position). Then you can combine those and output the result as a masked array or list:

def combine_grades(passed, failed):
    # Put index:value for each list into a dictionary
    pass_dict = {list(passed).index(i):i for i in list(passed) if i != '--'}
    fail_dict = {list(failed).index(i):i for i in list(failed) if i != '--'}

    # Combine those dictionaries
    full_dict = {**pass_dict, **fail_dict}

    # Return a masked array (if that's what you want. Otherwise, just return the list)
    return ma.array([full_dict[i] for  i in sorted(full_dict)])

Then call:

combine_grades(grades_passed, grades_not_passed)

Output:

masked_array(data=[ 90,  54,  88,  62,  34, 100],
             mask=False,
       fill_value=999999)

Upvotes: 0

Related Questions