def predict(self, x_, nlv_=None):
x = mx.NewMatrix(x_)
pscores_ = mx.initMatrix()
py_ = mx.initMatrix()
nlv = None
if nlv_ is None:
nlv = self.nlv
else:
nlv = self.nlv
PLSYPredictorAllLV(x, self.mpls, pscores_, py_)
pscores = mx.MatrixToList(pscores_)
py = mx.MatrixToList(py_)
mx.DelMatrix(x)
del x
mx.DelMatrix(pscores_)
del pscores_
mx.DelMatrix(py_)
del py_
return py, pscores
def reconstruct_original_matrix(self, npc_=None, scores_=None):
npc = None
if npc_ is None:
npc = self.npc
else:
npc = npc_
scores = None
if scores_ is None:
scores = self.mpca[0].scores
else:
scores = scores_
ivals = mx.initMatrix()
PCAIndVarPredictor(scores, self.mpca[0].loadings,
self.mpca[0].colaverage, self.mpca[0].colscaling,
npc, ivals)
omx = mx.MatrixToList(ivals)
mx.DelMatrix(ivals)
del ivals
return omx
def get_weights(self):
return mx.MatrixToList(self.mpls[0].xweights)
def get_qloadings(self):
return mx.MatrixToList(self.mpls[0].yloadings)
def get_uscores(self):
return mx.MatrixToList(self.mpls[0].yscores)
def get_loadings(self):
return mx.MatrixToList(self.mpca[0].loadings)
def get_scores(self):
return mx.MatrixToList(self.mpca[0].scores)