changed the title of plots

goeckslab · Aug 1, 2024 · bcda55c · bcda55c
1 parent ca79cf5
commit bcda55c
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 11 deletions.
diff --git a/tools/feature_importance.py b/tools/feature_importance.py
@@ -20,10 +20,10 @@ def __init__(self, task_type, output_dir, data_path=None, data=None,  target_col
             self.data = data
         else:
             self.target_col = target_col
-            self.task_type = task_type
             self.data = pd.read_csv(data_path, sep=None, engine='python')
             self.data.columns = self.data.columns.str.replace('.', '_')
             self.data = self.data.fillna(self.data.median(numeric_only=True))
+        self.task_type = task_type
         self.target = self.data.columns[int(target_col) - 1]
         self.exp = ClassificationExperiment() if task_type == 'classification' else RegressionExperiment()
         self.plots = {}
@@ -96,21 +96,26 @@ def generate_html_report(self, coef_html):
             encoded_image = self.encode_image_to_base64(plot_path)
             plots_html += f"""
             <div class="plot" id="{plot_name}">
-                <h3>{plot_name.replace('_', ' ').capitalize()}</h3>
+                <h2>Feature importance analysis from a trained Random Forest</h2>
+                <h3>{'Use gini impurity for calculating feature importance for classification' 
+                     'and Variance Reduction for regression'
+                  if plot_name == 'tree_importance' 
+                  else 'SHAP Summary from a trained lightgbm'}</h3>
                 <img src="data:image/png;base64,{encoded_image}" alt="{plot_name}">
             </div>
             """
 
         # Generate HTML content with tabs
         html_content = f"""
-                <h1>PyCaret Feature Importance Report</h1>
+            <h1>PyCaret Feature Importance Report</h1>
 
-                <div id="coefficients" class="tabcontent">
-                    <h2>Coefficients (based on a trained Logistic Regression Model)</h2>
-                    <div>{coef_html}</div>
-                </div>
-                {plots_html}
-            """
+            <div id="coefficients" class="tabcontent">
+                <h2>Coefficients (based on a trained 
+                {'Logistic Regression' if self.task_type == 'classification' else 'Linear Regression'} Model)</h2>
+                <div>{coef_html}</div>
+            </div>
+            {plots_html}
+        """
 
         return html_content
 

diff --git a/tools/pycaret_train.xml b/tools/pycaret_train.xml
@@ -1,5 +1,5 @@
 <tool id="pycaret_tool" name="PyCaret Model Comparison" version="@VERSION@" profile="@PROFILE@">
-    <description>Compare different machine learning models on a dataset using PyCaret.</description>
+    <description>Compare different machine learning models on a dataset using PyCaret. Do feature analysis using LR, Random Forest and LightGBM. </description>
     <macros>
         <import>pycaret_macros.xml</import>
     </macros>
@@ -164,7 +164,6 @@
     </tests>
     <help>
         This tool uses PyCaret to train and evaluate machine learning models.
-        Ensure that the Conda environment specified in the requirements is correctly set up.
     </help>
     <expand macro="macro_citations" />
 </tool>