def _fetch_file(self, input_path, **param):
if self.model == None:
raise 'Please Train a model before predicting some thing'
self.update_param(param)
obj_dt = DataTransform(input_path, param)
obj_dt.scan_data_type()
x, y = obj_dt.fetch_data()
return (x, y)
def train(self,input_path,**param):
self.update_param(param)
obj_dt = DataTransform(input_path,param)
obj_dt.scan_data_type()
x,y = obj_dt.fetch_data()
clf = linear_model.Lasso(alpha=0.01)
model = clf.fit(x,y)
self.model = model
return model
def train(self,input_path,**param):
self.update_param(param)
obj_dt = DataTransform(input_path,param)
obj_dt.scan_data_type()
x,y = obj_dt.fetch_data()
clf = svm.SVC()
model = clf.fit(x,y)
self.model = model
return model
def train(self,input_path,**param):
self.update_param(param)
if not self.int_dimension == None:
param['n_features'] = self.int_dimension
obj_dt = DataTransform(input_path,param)
obj_dt.scan_data_type()
x,y = obj_dt.fetch_data()
clf = tree.DecisionTreeClassifier()
model = clf.fit(x,y)
self.model = model
return model