script cleaning

src/pages/3-prediction.py

@@ -187,12 +187,6 @@ with c1:
 # Load parameters
 st.subheader("I - Spectral data preprocessing & visualization", divider="blue")
 if not pred_data.empty:  # Load the model with joblib
-    from utils.data_handling import signal_preprocess
-
-    preprocessed = signal_preprocess(
-        np.array(pred_data.loc[:, system_data["data"]["wls"]]),
-        tune=system_data["spec-preprocessing"],
-    )
     try:
         from utils.data_handling import signal_preprocess
 
@@ -208,20 +202,16 @@ if not pred_data.empty:  # Load the model with joblib
 if not preprocessed.empty:
     c2, c3 = st.columns([2, 1])
     with c2:
-        rawspectraplot = plot_spectra(
-            pred_data,
-            color=None,
-            cmap=None,
-            xunits="Wavelength/Wavenumber",
-            yunits="Signal intensity",
-        )
-        prepspectraplot = plot_spectra(
-            preprocessed,
-            color=None,
-            cmap=None,
-            xunits="Wavelength/Wavenumber",
-            yunits="Signal intensity",
-        )
+        rawspectraplot, prepspectraplot = [
+            plot_spectra(
+                i,
+                color=None,
+                cmap=None,
+                xunits="Wavelength/Wavenumber",
+                yunits="Signal intensity",
+            )
+            for i in (pred_data, preprocessed)
+        ]
 
         st.write("Raw spectra")
         st.pyplot(rawspectraplot)
@@ -280,6 +270,7 @@ if not preprocessed.empty:
                 )
             )
             return fig
+
         pca = calpred_pca(
             cal=system_data["data"]["raw-spectra"]
             .iloc[system_data["data"]["idx"]["train"], :]
@@ -299,7 +290,7 @@ if not preprocessed.empty:
 
     disable2 = False if not pred_data.empty else True
     pred_button = st.button(
-        "Predict " + str(system_data["data"]["target"]['name']) + " values",
+        "Predict " + str(system_data["data"]["target"]["name"]) + " values",
         type="primary",
         disabled=disable2,
         use_container_width=False,
@@ -308,30 +299,31 @@ if not preprocessed.empty:
         st.session_state["predict"] = True
 
     if not preprocessed.empty and st.session_state["predict"]:
-        model = system_data['model']["model_"]
-        if system_data['model']["model_type"] in ["PLS", "TPE-iPLS"]:
-            nvar = system_data['model']["model_"].n_features_in_
-        elif system_data['model']["model_type"] == "LW-PLS":
+        model = system_data["model"]["model_"]
+        if system_data["model"]["model_type"] in ["PLS", "TPE-iPLS"]:
+            nvar = system_data["model"]["model_"].n_features_in_
+        elif system_data["model"]["model_type"] == "LW-PLS":
             nvar = system_data["data"]["raw-spectra"].shape[1]
 
-        if system_data['model']["selected-wls"] is None:
+        if system_data["model"]["selected-wls"] is None:
             preprocesseddf = preprocessed
         else:
-            preprocesseddf = preprocessed.loc[:, system_data['model']["selected-wls"]]
+            preprocesseddf = preprocessed.loc[:, system_data["model"]["selected-wls"]]
 
         if not preprocesseddf.empty:
-            match system_data['model']["model_type"]:
+            match system_data["model"]["model_type"]:
                 case "PLS" | "TPE-iPLS":
                     if preprocesseddf.shape[1] == nvar:
                         result = DataFrame(
-                                system_data['model']["model_"].predict(
-                                    preprocesseddf.to_numpy()
-                                ),
-                                index=preprocesseddf.index,
-                                columns=["Results"])
+                            system_data["model"]["model_"].predict(
+                                preprocesseddf.to_numpy()
+                            ),
+                            index=preprocesseddf.index,
+                            columns=["Results"],
+                        )
                         try:
                             result = DataFrame(
-                                system_data['model']["model_"].predict(
+                                system_data["model"]["model_"].predict(
                                     preprocesseddf.to_numpy()
                                 ),
                                 index=preprocesseddf.index,
@@ -350,10 +342,11 @@ if not preprocessed.empty:
                     try:
 
                         spectra = signal_preprocess(
-                                np.array(system_data["data"]["raw-spectra"])[
-                                system_data["data"]["idx"]['train'], :],
-                                tune=system_data["spec-preprocessing"],
-                            )
+                            np.array(system_data["data"]["raw-spectra"])[
+                                system_data["data"]["idx"]["train"], :
+                            ],
+                            tune=system_data["spec-preprocessing"],
+                        )
 
                         from utils.lwplsr_julia_converted import lwpls
 
@@ -362,10 +355,11 @@ if not preprocessed.empty:
                                 Xtrain=np.array(spectra),
                                 Xtest=np.array(preprocessed),
                                 ytrain=np.array(
-                                    system_data["data"]["target"]['target'].iloc[
-                                        system_data["data"]["idx"]['train']
+                                    system_data["data"]["target"]["target"].iloc[
+                                        system_data["data"]["idx"]["train"]
                                     ]
-                                ),**system_data['model']['model_']
+                                ),
+                                **system_data["model"]["model_"],
                             ),
                             index=preprocessed.index,
                         )
@@ -432,8 +426,8 @@ if not result.empty:
         st.dataframe(resultT.T)
     with c5:
         st.info("The performance of the model used for prediction making")
-        st.table(system_data['model']["performance"])
-        st.pyplot(system_data['model']["measvspred"])
+        st.table(system_data["model"]["performance"])
+        st.pyplot(system_data["model"]["measvspred"])
         from utils.miscellaneous import desc_stats
 
         st.info("descriptive statistics for the model output")