Overview
- The numpy.unpackbits() unpacks an ndarray of numpy.uint8 integers into an ndarray of zeros and ones.
- Note that the data type of the ndarray is not numpy.bool. It remains as numpy.uint8.
- The numpy.unpackbits() does the reverse of numpy.packbits().

Example 1
# Example Python program that unpacks an ndarray of numpy.uint8 instances
# into an ndarray of bits
import numpy
# Create an ndarray. Elements are of type numpy.uint8
intArray = numpy.ndarray(shape = (2, 2),
dtype = numpy.uint8)
# Fill the ndarray with values
intArray[0][0] = 10
intArray[0][1] = 15
intArray[1][0] = 20
intArray[1][1] = 25
# Get the binary bits of the integer as an ndarray
bitArray = numpy.unpackbits(intArray, axis = 1)
print("Input array:")
print(intArray)
print("Bit array when axis=1:")
print(bitArray)
print(bitArray.dtype)
bitArray = numpy.unpackbits(intArray, axis = 0)
print("Bit array when axis=0:")
print(bitArray)Output
Input array:
[[10 15]
[20 25]]
Bit array when axis=1:
[[0 0 0 0 1 0 1 0 0 0 0 0 1 1 1 1]
[0 0 0 1 0 1 0 0 0 0 0 1 1 0 0 1]]
uint8
Bit array when axis=0:
[[0 0]
[0 0]
[0 0]
[0 0]
[1 1]
[0 1]
[1 1]
[0 1]
[0 0]
[0 0]
[0 0]
[1 1]
[0 1]
[1 0]
[0 0]
[0 1]]Example 2
# Example Python program that unpacks an ndarray
# of numpy.uint8 types in little endian bit order
import numpy
intVals = numpy.ndarray(shape = (2, 2),
dtype = numpy.uint8)
intVals[0][0] = 1
intVals[0][1] = 120
intVals[1][0] = 128
intVals[1][1] = 255
unpacked = numpy.unpackbits(intVals, bitorder = 'little')
print("An ndarray of numpy.uint8 elements:")
print(intVals)
print("Unpacked ndarray - least significant bits are placed first:")
print(unpacked)Output
An ndarray of numpy.uint8 elements:
[[ 1 120]
[128 255]]
Unpacked ndarray - least significant bits are placed first:
[1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1]