numpy 차원수 늘리기 / 줄이기

워낙 기초가 안되어 있다보니,

 

numpy 변수의 차원을 늘리는 것조차 급하면 생각이 안나서 기록해둠

 

print('test_images[0].shape : ', test_images[0].shape)

expand_dims_val0 = np.expand_dims(test_images[0], axis=0)
print('expand_dims_val0.shape : ', expand_dims_val0.shape)

expand_dims_val1 = np.expand_dims(test_images[0], axis=1)
print('expand_dims_val1.shape : ', expand_dims_val1.shape)

expand_dims_val2 = np.expand_dims(test_images[0], axis=2)
print('expand_dims_val1.shape : ', expand_dims_val2.shape)

print('squeeze_val0.shape : ', np.squeeze(expand_dims_val0).shape)
print('squeeze_val1.shape : ', np.squeeze(expand_dims_val1).shape)
print('squeeze_val2.shape : ', np.squeeze(expand_dims_val2).shape)

 

[결과]

test_images[0].shape :  (28, 28)
expand_dims_val0.shape :  (1, 28, 28)
expand_dims_val1.shape :  (28, 1, 28)
expand_dims_val1.shape :  (28, 28, 1)
squeeze_val0.shape :  (28, 28)
squeeze_val1.shape :  (28, 28)
squeeze_val2.shape :  (28, 28)