[FIXED] how to use values of a pandas DataFrame as numpy array index

Issue

I have a pandas dataFrame like this.

The X, Y, Z are they (x,y,z) coordinates that represent a point inside a cube of side-length 255.

I want to create a numpy array/dataFrame from this, whose index will be the (x,y,z) coordinates and value is the intensity.

the output should be

data[133,55,250] = 8
data[133,61,254] = 21
...

I tried something like this

data = np.zeros((255,255,255), dtype=np.int)
index = np.array((temp['X'], temp['Y'], temp['Z']))

but returned index is a (3,15) array.

I want a index where

data[index] = intensity

will give me my result.

Solution

Instead of

index = np.array((temp['X'], temp['Y'], temp['Z']))

you could use integer array indexing to perform the assignment:

data = np.zeros((256, 256, 256), dtype=np.int)
data[temp['X'], temp['Y'], temp['Z']] = temp['intensity']

Answered By - unutbu