Design（设计方案）

• Describe/Explain the general idea of the design. Pseudo-code is allowed

TASK 1

for

mindspore_xai.visual.cv
normalize_saliency()
saliency_to_rgba()
saliency_to_image()

mindspore_xai.explainer
PseudoLinearCoef()
PseudoLinearCoef.normalize()
PseudoLinearCoef.plot()

add checkings for every input arguments (at the beginning of the function):

1. none checking (if which cannot be none)
2. type checking
3. element type checking (if argument is a list tuple tensor or numpy)
4. value range checking
   raise ValueError immediately if any checking failed

TASK 2

update the APIs, add optional/可选 to the arguments with default value

TASK 3

1. update PLC API doc and checking code, predictor can accept both ms.float32 and ms.float64 as input/output
2. update PLC API doc and checking code, input argument: features dtype can be ms.float32 and ms.float64