From 2d045dd9eecc05d93eb1d80d7da740cb194d0069 Mon Sep 17 00:00:00 2001
From: DIANE <abderrahim.diane@cefe.cnrs.fr>
Date: Thu, 4 Apr 2024 09:59:48 +0200
Subject: [PATCH] PLSR pinard update
---
Class_Mod/PLSR_.py | 113 +++++++++++++++++++++++----------------------
1 file changed, 59 insertions(+), 54 deletions(-)
diff --git a/Class_Mod/PLSR_.py b/Class_Mod/PLSR_.py
index c76718a..4070347 100644
--- a/Class_Mod/PLSR_.py
+++ b/Class_Mod/PLSR_.py
@@ -1,58 +1,63 @@
from Packages import *
from Class_Mod.Miscellaneous import *
+from Class_Mod.Regression_metrics import metrics
-# create model module with PINARD
-def model_PLSR(xcal_csv, ycal_csv, sep, hdr, rd_seed):
- np.random.seed(rd_seed)
- # hdr var correspond to column header True or False in the CSV
- if hdr == 'yes':
- col = 0
- else:
- col = False
- # loading the csv
- x, y = utils.load_csv(xcal_csv, ycal_csv, autoremove_na=True, sep=sep, x_hdr=0, y_hdr=0, x_index_col=col, y_index_col=col)
- # Split data into training and test sets using the kennard_stone method and correlation metric, 25% of data is used for testing
- train_index, test_index = train_test_split_idx(x, y=y, method="kennard_stone", metric="correlation", test_size=0.25, random_state=rd_seed)
- # Assign data to training and test sets
- X_train, y_train, X_test, y_test = x[train_index], y[train_index], x[test_index], y[test_index]
- st.write("Size of train and test sets: train " + str(X_train.shape) + ' ' + str(y_train.shape) + ' / test ' + str(X_test.shape) + ' ' + str(y_test.shape))
- # Declare preprocessing pipeline
- svgolay = [ ('_sg1',pp.SavitzkyGolay()),
- ('_sg2',pp.SavitzkyGolay()) # nested pipeline to perform the Savitzky-Golay method twice for 2nd order preprocessing
- ]
- preprocessing = [ ('id', pp.IdentityTransformer()), # Identity transformer, no change to the data
- ('savgol', pp.SavitzkyGolay()), # Savitzky-Golay smoothing filter
- ('derivate', pp.Derivate()), # Calculate the first derivative of the data
- ('SVG', FeatureUnion(svgolay))
- # Pipeline([('_sg1',pp.SavitzkyGolay()),('_sg2',pp.SavitzkyGolay())]) # nested pipeline to perform the Savitzky-Golay method twice for 2nd order preprocessing
- ]
- # Declare complete pipeline
- pipeline = Pipeline([
- ('scaler', MinMaxScaler()), # scaling the data
- ('preprocessing', FeatureUnion(preprocessing)), # preprocessing
- ('PLS', PLSRegression()) # regressor
- ])
- # Estimator including y values scaling
- estimator = TransformedTargetRegressor(regressor = pipeline, transformer = MinMaxScaler())
- # Training
- trained = estimator.fit(X_train, y_train)
- # fit scores
- st.write("fit scores / R²: " + str(estimator.score(X_test,y_test)))
- # Predictions on test set
- Y_preds = estimator.predict(X_test) # make predictions on test data and assign to Y_preds variable
- ################################################################################################################
- met= {"MAE: ":[5],
- "MSE: ":[5],
- "MSE: ":[8]}
- met = pd.DataFrame(met).T
- ################################################################################################################
- st.table(met)
- st.write("MAE: " + str(mean_absolute_error(y_test, Y_preds)))
- st.write("MSE: " + str(mean_squared_error(y_test, Y_preds)))
- st.write("MAPE: " + str(mean_absolute_percentage_error(y_test, Y_preds)))
-
- # Cross-Validate the model
- CV_model(estimator, X_train, y_train, 3)
-
- return (trained)
\ No newline at end of file
+class PinardPlsr:
+ def __init__(self, x_train, y_train, x_test, y_test):
+ self.x_train = x_train
+ self.x_test = x_test
+ self.y_train = y_train
+ self.y_test = y_test
+
+ # create model module with PINARD
+ # Declare preprocessing pipeline
+ svgolay = [ ('_sg1',pp.SavitzkyGolay()),
+ ('_sg2',pp.SavitzkyGolay()) # nested pipeline to perform the Savitzky-Golay method twice for 2nd order preprocessing
+ ]
+ preprocessing = [ ('id', pp.IdentityTransformer()), # Identity transformer, no change to the data
+ ('savgol', pp.SavitzkyGolay()), # Savitzky-Golay smoothing filter
+ ('derivate', pp.Derivate()), # Calculate the first derivative of the data
+ ('SVG', FeatureUnion(svgolay))
+ # Pipeline([('_sg1',pp.SavitzkyGolay()),('_sg2',pp.SavitzkyGolay())]) # nested pipeline to perform the Savitzky-Golay method twice for 2nd order preprocessing
+ ]
+ # Declare complete pipeline
+ pipeline = Pipeline([
+ ('scaler', MinMaxScaler()), # scaling the data
+ ('preprocessing', FeatureUnion(preprocessing)), # preprocessing
+ ('PLS', PLSRegression())])
+ # Estimator including y values scaling
+ estimator = TransformedTargetRegressor(regressor = pipeline, transformer = MinMaxScaler())
+ # Training
+ self.trained = estimator.fit(self.x_train, self.y_train)
+
+
+ # fit scores
+ # Predictions on test set
+ self.yc = pd.DataFrame(self.trained.predict(self.x_train)) # make predictions on test data and assign to Y_preds variable
+ self.ycv = pd.DataFrame(cross_val_predict(self.trained, self.x_train, self.y_train, cv = 3)) # make predictions on test data and assign to Y_preds variable
+ self.yt = pd.DataFrame(self.trained.predict(self.x_test)) # make predictions on test data and assign to Y_preds variable
+
+ ################################################################################################################
+
+
+
+ ################################################################################################################
+
+ @property
+ def model_(self):
+ return self.trained
+
+ @property
+ def metrics_(self):
+ metc = metrics(train=(self.y_train, self.yc))
+ metcv = metrics(train=(self.y_train, self.ycv))
+ mett = metrics( train=(self.y_test, self.yt))
+ met = pd.concat([metc, metcv, mett], axis = 0)
+ met.index = ['calib','cv','test']
+ return met
+
+ @property
+ def pred_data_(self):
+
+ return self.yc, self.ycv, self.yt
\ No newline at end of file
--
GitLab