Added simulation and explainer

src/manocca_package/manocca.py

@@ -46,7 +46,7 @@ class MANOCCA:
             self.prodV_red = prodV_red
 
         else : # If not we preprocess the data and compute prodV and prodV_red
-            self.prodV = self.get_prodV_para_wrap(self.output, n_jobs = self.n_jobs)
+            self.prodV = self.get_prodV_para_wrap(scale(self.output), n_jobs = self.n_jobs)
             self.prodV = scale(self.prodV)
             self.prodV_red, self.pca = self.get_prodV_red(self.prodV, self.n_comp, return_pca = True) # prodV_red is scaled within the get_prodV_red function
 
@@ -80,19 +80,22 @@ class MANOCCA:
             return scale(tmp)
 
 
-    def test(self, n_comp = None):
+    def test(self, var = None, n_comp = None, min_comp = 0):
         if n_comp is None :
             n_comp = self.prodV_red.shape[1]
-        n_preds = self.predictors.shape[1]
+        if var is None:
+            var = self.predictors
+
+        n_preds = var.shape[1]
         p = np.empty(n_preds)
         for i_pred in range(n_preds):
-            nan_mask = np.isnan(self.predictors[:,i_pred])
+            nan_mask = np.isnan(var[:,i_pred])
             if self.covariates is not None :
                 print("With covariates")
-                p[i_pred] = custom_manova(self.prodV_red[~nan_mask,:n_comp], scale(self.predictors[~nan_mask,i_pred].reshape(-1,1)), scale(self.covariates[~nan_mask,:]))
+                p[i_pred] = custom_manova(self.prodV_red[~nan_mask,min_comp:n_comp], scale(var[~nan_mask,i_pred].reshape(-1,1)), scale(self.covariates[~nan_mask,:]))
             else :
                 print("Without covariates")
-                p[i_pred] = custom_manova(self.prodV_red[~nan_mask,:n_comp], scale(self.predictors[~nan_mask,i_pred].reshape(-1,1)))
+                p[i_pred] = custom_manova(self.prodV_red[~nan_mask,min_comp:n_comp], scale(var[~nan_mask,i_pred].reshape(-1,1)))
         self.p = p
 
 

src/manocca_package/tools/explainer.py

+from manocca import MANOCCA
+from manova import MANOVA
+
+
+
+class Explainer :
+	def __init__(self, model):
+		self.model = model
+
+
+	def pc_
\ No newline at end of file

src/manocca_package/tools/simulation.py

+from manocca import MANOCCA
+from manova import MANOVA
+
+import tools.preprocessing_tools as pt
+
+class Simu :
+    def __init__(self,  methods, output, predictors, covariates=None, cols_output = None,
+                 cols_predictors = None, cols_covariates = None, L_preproc = [], prodV_red=None, n_comp = None, n_jobs = 1):
+
+        self.methods = methods
+
+            ### Initializing 
+        self.output = output
+        self.cols_output = cols_output
+        self.output, self.cols_output = pt._extract_cols(self.output, self.cols_output)
+
+        self.predictors = predictors
+        self.cols_predictors = cols_predictors
+        self.predictors, self.cols_predictors = pt._extract_cols(self.predictors, self.cols_predictors)
+
+        self.covariates = covariates
+        self.cols_covariates = cols_covariates
+        if covariates is not None :
+            self.covariates, self.cols_covariates = pt._extract_cols(self.covariates, self.cols_covariates) 
+
+        self.prodV_red = prodV_red
+
+        self.n_comp = n_comp
+        self.n_jobs = n_jobs
+
+        if len(L_preproc)>0:
+            self.output = pt.pipeline(self.output, L_pipe = L_preproc)
+
+        self.L_models = []
+        self._get_model()
+
+        ### To fill later ###
+        self.p = np.empty((self.predictors.shape[0],0))
+
+    def _get_model(self):
+        for m in self.methods :
+            if m == 'MANOCCA':
+                self.L_models += [MANOCCA(self.output, self.predictors, self.covariates, self.cols_output, self.cols_predictors, self.cols_covariates, self.prodV_red, self.n_comp, self.n_jobs)]
+            elif m == 'MANOVA':
+                self.L_models += [MANOVA(self.output, self.predictors, self.covariates, self.cols_output, self.cols_predictors, self.cols_covariates)]
+    
+    def run_simu(self, n_comp):
+        for m in self.L_models :
+            m.test()
+            self.p = np.hstack([self.p, m.p])
+
+