torch.distributions.uniform.Uniform.cdf() can return negative values or values above one #7532
Description
Issue description
torch.distributions.uniform.Uniform.cdf() can return negative values or values above one because it does not do a range check like log_prob() in the same class does.
Relevant code in cdf():
pytorch/torch/distributions/uniform.py
Lines 70 to 74 in 825c3ca
Range checking in log_prob():
pytorch/torch/distributions/uniform.py
Lines 63 to 68 in 825c3ca
I can provide a pull request based on the range checking as done in the log_prob() method of the same class (although log_prob() throws an exception with scalar arguments because of these range checks while cdf() accepts scalar arguments).
Code example
import torch
from torch.distributions.uniform import Uniform
uniform = Uniform(1,2)
print uniform.cdf(0), uniform.cdf(5)
will give
tensor(-1.) tensor(4.)
System Info
Collecting environment information...
PyTorch version: 0.4.0
Is debug build: No
CUDA used to build PyTorch: Could not collect
OS: Mac OSX 10.13.4
GCC version: Could not collect
CMake version: version 3.11.0
Python version: 2.7
Is CUDA available: No
CUDA runtime version: 9.1.128
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Could not collect
Versions of relevant libraries:
[pip] numpy (1.14.2)
[pip] root-numpy (4.4.0)
[pip] torch (0.4.0)
[pip] torchvision (0.2.1)
[conda] Could not collect