How to calculate the average of 2 different numpy array pairs of bunch of points in list?-CodePudding

I have a numpy array pair list

[[214,295], [215, 294], [229, 226], [229, 227]]

After calculating the average of the bunch of points using the Z score, the result I have is

[[222.0, 260.5], [214.0, 295.0], [229.0, 226.0]]

Expected result [[average of [214, 295], [215, 294]] , [average of [229, 226], [229, 227]]

It should always return 2 points instead of more than 3 points. Do I need to calculate a new Z score separately?

Another example that I can bring up for discussion

[[95, 132], [96, 132], [94, 133], [134, 239], [95, 131]]

Current output

[[ 95. 132.], [134. 239.]]

If the data has many points bunched around the [134, 239] point, I would like to have a more robust way to split the two main bunches.

import numpy as np
from scipy import stats

tempList = np.array([[214,295], [215, 294],[229, 226], [229, 227]])

z= stats.zscore(tempList, axis=0)
z = list([abs(x)<1 and abs(y)<1 for x,y in z])

newList = tempList[[not x for x in z]]

tempList = tempList[z]

newList = np.concatenate([[tempList.mean(axis=0)], newList])
print(newList)

CodePudding user response：

Well, it's easy to do what you describe just by reshaping the array:

import numpy as np
tempList = np.array([[214,295], [215, 294],[229, 226], [229, 227]])
tempList = tempList.reshape( (-1,2,2) )
print(tempList)
print("---")
print( tempList.mean( axis=1 ) )

Output:

[[[214 295]
  [215 294]]

 [[229 226]
  [229 227]]]
---
[[214.5 294.5]
 [229.  226.5]]