Some .....

fluents/lib/blmfuncs.py

@@ -171,7 +171,7 @@ class PCA(Model):
         self.make_model(**options.make_model_options())
         if options['calc_conf']:
             self.confidence(**options.confidence_options())
-            
+        
         out = [self.as_dataset(p) for p in options['out_data']]
         for plt in self.get_out_plots(options):
             out.append(plt)
@@ -206,6 +206,7 @@ class PLS(Model):
         else:
             val_engine = pls_val
         if self._options['calc_cv']==True:
+            print "Doing RMSEP calc ..."
             rmsep, aopt = val_engine(self.model['E0'], self.model['F0'],
                                      self._options['amax'], self._options['n_sets'])
             self.model['rmsep'] = rmsep[:,:-1]
@@ -221,6 +222,7 @@ class PLS(Model):
         """
         aopt = self.model['aopt']
         if self._options['calc_conf']:
+            print "Doing Tsq"
             jk_segments = pls_jkW(self.model['E0'], self.model['F0'],
                                   aopt, n_sets)
             Wcal = self.model['W'][:,:aopt]
@@ -238,6 +240,7 @@ class PLS(Model):
         """Estimates cut off on significant vars by controlling fdr."""
 
         if self._options['calc_qvals']==True:
+            print "Doing Qvals"
             qvals = pls_qvals(a, b,
                               aopt=None,
                               alpha=reg,
@@ -253,6 +256,7 @@ class PLS(Model):
     def make_model(self, a, b, amax, scale, mode, engine):
         """Make model on amax components.
         """
+        print "MAking model"
         dat = engine(a, b, amax, scale, mode)
         self.model.update(dat)
         
@@ -330,18 +334,7 @@ class PLS(Model):
         var_y, exp_var_y = variances(self.model['F0'], self.model['T'], self.model['Q'])
         self.model['var_y'] = var_y
         self.model['exp_var_y'] = exp_var_y
-        
+                    
-        if options['calc_corrloads']:
-            corr_load = scipy.empty_like(self.model['P'].copy())
-            T = self.model['T']
-            X = self._data['X']
-            # For each variable/attribute in original matrix (not meancentered)
-            for i,score in enumerate(T.T):
-                for j, profile in enumerate(X.T):
-                    corrs = scipy.corrcoef(score, profile)
-                    corr_load[j,i] = corrs[0,1]
-            self.model['CP'] = corr_load
-            
         if options['calc_conf']:
             self.confidence(**options.confidence_options())
 
@@ -598,13 +591,12 @@ class PlsOptions(Options):
         opt['center'] = True
         opt['center_mth'] = mat_center
         opt['scale'] = 'scores'
-
+        
-        opt['calc_corrloads'] = True
         opt['calc_conf'] = False
-        opt['n_sets'] = 5
+        opt['n_sets'] = 7
         opt['strict'] = True
         opt['p_center'] = 'med'
-        opt['alpha'] = .8
+        opt['alpha'] = .2
         opt['cov_center'] = 'med'
         opt['crot'] = True
 
@@ -625,7 +617,7 @@ class PlsOptions(Options):
                             (blmplots.LineViewXc, 'Line view', True),
                             (blmplots.PredictionErrorPlot, 'Residual Error', False),
                             (blmplots.RMSEPPlot, 'RMSEP', False),
-                            (blmplots.PlsCorrelationLoadingPlot, 'Corr. loadings', True)
+                            (blmplots.PlsCorrelationLoadingPlot, 'Corr. loadings', False)
                             ]
         
         opt['out_data'] = ['T','P', 'p_tsq']
@@ -673,10 +665,10 @@ class LplsOptions(Options):
         opt['mode'] = 'normal' # how much info to calculate
         opt['amax'] = 10
         opt['aopt'] = 4
-        opt['xz_alpha'] = .3
+        opt['xz_alpha'] = 0.5
         opt['auto_aopt'] = False
         opt['center'] = True
-        opt['center_mth'] = [2, 0, 1]
+        opt['center_mth'] = [2, 2, 1]
         opt['scale'] = 'scores'
         opt['calc_conf'] = False
         opt['n_sets'] = 75