-3

array([1500, 1520, 1540, 1590, 1590, 1600, 1600, 1560, 1560, 1560, 1580, 1520, 1460, 1510, 1520, 1320, 1320, 1300, 1300, 1320, 1320, 1320, 1320, 1300, 1300, 1340, 1300, 1300, 1300, 1400, 1480, 1360, 1420, 1480, 1580, 1530, 1500, 1480, 1480, 1480, 1460, 1540, 1490, 1480, 1480, 1520, 1500, 1460, 1480, 1480, 1500, 1500, 1600, 1540, 1480, 1460, 1560, 1600, 1560, 1600, 1600, 1600, 1620, 1600, 1580, 1600, 1700, 1620, 1620, 1620, 1700, 1700, 1680, 1640, 1620, 1670, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1670, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1600, 1680, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64)

anonymous
  • 13
  • 3

3 Answers3

2

You can filter the array

arr = np.array([41, 42, 43, 44, 0, 0,0])
arr.mean() # 24.2857
arr[arr != 0].mean() # 42.5

edit: (!= to include negative numbers thanks to @tdelaney)

Chinny84
  • 956
  • 6
  • 16
1

You can do that with list comprehension and numpy:

import numpy as np
a = np.array([1500, 1520, 1540, 1590, 1590, 1600, 1600, 1560, 1560, 1560, 1580,
              1520, 1460, 1510, 1520, 1320, 1320, 1300, 1300, 1320, 1320, 1320, 1320,
              1300, 1300, 1340, 1300, 1300, 1300, 1400, 1480, 1360, 1420, 1480,
              1580, 1530, 1500, 1480, 1480, 1480, 1460, 1540, 1490, 1480, 1480,
              1520, 1500, 1460, 1480, 1480, 1500, 1500, 1600, 1540, 1480, 1460,
              1560, 1600, 1560, 1600, 1600, 1600, 1620, 1600, 1580, 1600, 1700,
              1620, 1620, 1620, 1700, 1700, 1680, 1640, 1620, 1670, 0, 0, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1670, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1600, 1680, 0, 0, 0, 0,
              0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
              0, 0, 0], dtype=np.int64)

np.mean([i for i in a if i])

Output:

1508.4810126582279
Nin17
  • 2,821
  • 2
  • 4
  • 14
  • 1
    This is an expensive operation - each element of the raw numpy array needs to be converted to a python object. Better to keep it as an numpy arrray and use its operators. – tdelaney Apr 27 '22 at 22:27
1

Here's a simple approach is to just make a copy of the array without the zeros, then calculate the average, this way:

import numpy as np
new_array = array[array!=0] #Create a copy of the array excluding the zeros
avg = np.average(new_array)

And here's another idea that doesn't use any of numpy methods:

count=0
total=0
for number in array:
    if number != 0:
        total+=number #adds the non-zero values together
        count+=1      #counts the non-zero values
avg= total/count
Heidi Hussein
  • 13
  • 5