birg / pyopls Goto Github PK

Good afternoon,

When running the example code for Validator I am unable to get the OPLSValidator to fit without throwing the ValueError: Only one class present in y_true. ROC AUC score is not defined in that case.

Not sure if this has something to do with my setup or if others are having this issue as well.

Thanks!
Scott

Hi, does this repo also work well with multi-class classification instead of binary classification?
Thank you,

Hi I have a quick question;I will appreciate if you could explain that;
I see that in OPLS 39 components are used and in pls one component is passed in the code below;
#opls = OPLS(39)
#Z = opls.fit_transform(spectra, target)
#pls = PLSRegression(1)

Q1: What is the reason for using one component in "PLSRegression(1)" and how do you decide the number of OPLS components OPLS(39);

Q2: In Permutation test; what is this argument k=-1 referring to? and where can we access the information about what numbers of components were best for this model after doing permutation test?
#validator = OPLSValidator(k=-1).fit(X, target)

I appreciate you time and response;
Thanks
Amnah

The validator scripts are not working on any datasets, seems to be an issue with the fit method and the subsequent LabelBinarizer call. This is the error when running the example code:

TypeError                                 Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_19292/3997691689.py in <module>
     10 X = df[[c for c in df.columns if c!='classification']]
     11 
---> 12 validator = OPLSValidator(k=-1).fit(X, target)
     13 
     14 Z = validator.opls_.transform(X)

C:\devel\anaconda3\lib\site-packages\pyopls\validation.py in fit(self, X, y, n_components, cv, pos_label, random_state, n_jobs, verbose, pre_dispatch)
    452 
    453         X = check_array(X, dtype=float, copy=True)
--> 454         y = self._check_target(y, pos_label)
    455 
    456         if not n_components:

C:\devel\anaconda3\lib\site-packages\pyopls\validation.py in _check_target(self, y, pos_label)
    223                              'Try binarizing with sklearn.preprocessing.LabelBinarizer.')
    224         if self.is_discrimination(y):
--> 225             y = self._process_binary_target(y, pos_label)
    226         else:
    227             self.binarizer_ = None

C:\devel\anaconda3\lib\site-packages\pyopls\validation.py in _process_binary_target(self, y, pos_label)
    211 
    212     def _process_binary_target(self, y, pos_label=None):
--> 213         self.binarizer_ = LabelBinarizer(-1, 1)
    214         self.binarizer_.fit(y)
    215         if pos_label is not None and self.binarizer_.transform([pos_label])[0] == -1:

TypeError: __init__() takes 1 positional argument but 3 were give
```n

Hi,

The x and y axes were mislabeled in the score plot, which should be:

plt.ylabel('t_ortho')
plt.xlabel('t')

Dong

Setting n_jobs in OPLSValidator.fit() does not change the number of CPUs used. I suspect this is an issue with the joblib commands but if i find a solution i will do a pull request