Closed akhilsinghal1234 closed 5 years ago
I see a fit
function is missing and optimizer
is not declared. I will suggest
fit function that takes
Xand
y` as input and returning output.optimizer
as an input in __init__
with default value as None
. In fit
function if self.optimizer
is None
, Adam
optimizer is used for the same.Also,
classifier.py
to classifiers.py
. Do the same with model.py
and rename it to models.py
.Could you put an example code of this class here? We'll put it in the docs later.
Hello @akhilsinghal1234! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:
There are currently no PEP 8 issues detected in this Pull Request. Cheers! :beers:
Instead of adding optimizer and activation in init, shouldn't they be in build model?
Oh yes. You are right. I guess then it's better to call that function compile
rather than build model
. (I saw another example where they used that name).
Also, looks at the comments by @pep8speaks above. It refreshes after every commit. Very useful!
One small change is left (see above). Also, since the dataset is small, data the files in a folder in data
module along with a loader function.
Please comment any changes needed. Added train.py and utils.py too.
Please change the location of dataset from mirapy/dataset/XRayBinary/Dataset/
to mirapy/dataset/XRayBinary/
and set is as the default value of data_directory
in load_dataset
.
Also, I have added load_dataset.py
in data
module. Remind me to move load_dataset
function in utils.py
to that script (maybe calling the function load_XRB_binary_dataset
.
Also, resolve add PEP8 issues mentioned above by @pep8speaks.
LGTM!
Unit testing to be done