Add comments to model.py.

aki.csv

@@ -311,7 +311,7 @@ n
 n
 n
 n
-n
+y
 n
 n
 n
@@ -611,7 +611,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 n
@@ -863,7 +863,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 n
@@ -1742,7 +1742,7 @@ n
 y
 n
 n
-n
+y
 n
 n
 n
@@ -2078,7 +2078,7 @@ y
 n
 n
 y
-y
+n
 y
 y
 n
@@ -2134,13 +2134,13 @@ n
 n
 n
 n
+n
 y
 y
-y
-n
 n
 n
 n
+y
 n
 y
 y
@@ -2241,7 +2241,7 @@ n
 n
 y
 n
-y
+n
 n
 n
 n
@@ -2259,7 +2259,7 @@ n
 n
 y
 n
-y
+n
 n
 n
 n
@@ -2752,7 +2752,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 n
@@ -3511,7 +3511,7 @@ y
 n
 n
 n
-n
+y
 n
 n
 n
@@ -3776,7 +3776,7 @@ n
 y
 y
 y
-n
+y
 y
 n
 y
@@ -3817,7 +3817,7 @@ n
 y
 n
 n
-y
+n
 y
 n
 n
@@ -3933,7 +3933,7 @@ n
 y
 y
 n
-y
+n
 n
 y
 n
@@ -3967,7 +3967,7 @@ n
 y
 n
 n
-n
+y
 n
 y
 y
@@ -4032,7 +4032,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 y
@@ -4358,7 +4358,7 @@ y
 n
 n
 n
-n
+y
 n
 y
 n
@@ -4760,7 +4760,7 @@ n
 n
 y
 y
-y
+n
 y
 n
 y
@@ -4935,7 +4935,7 @@ n
 n
 n
 n
-y
+n
 y
 n
 y
@@ -5204,7 +5204,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 y
@@ -5255,7 +5255,7 @@ n
 n
 n
 n
-y
+n
 n
 y
 y
@@ -5683,7 +5683,7 @@ n
 n
 n
 n
-y
+n
 n
 n
 n
@@ -5923,7 +5923,7 @@ n
 y
 n
 n
-y
+n
 y
 n
 n
@@ -6294,7 +6294,7 @@ y
 n
 n
 n
-n
+y
 n
 n
 n
@@ -6515,7 +6515,7 @@ y
 n
 n
 n
-y
+n
 n
 n
 n

model.py

@@ -3,37 +3,60 @@
 import argparse
 import csv
 import pandas as pd
-from sklearn.metrics import fbeta_score, classification_report
+from sklearn.metrics import fbeta_score
 from sklearn.preprocessing import StandardScaler
 from sklearn.linear_model import LogisticRegression
 
 
 
-def preprocess_data(input_file):
-    data = pd.read_csv(input_file)
-
+def preprocess_data(data):
+    """
+    Preprocesses the input data.
+    This function performs the following steps:
+    1. Reads the input data from a CSV file.
+    2. Computes the mean, minimum, maximum, and variance of all columns containing "creatinine_result".
+    3. Drops all columns containing "creatinine".
+    4. Fills missing values in the "age" column with the median age.
+    5. Maps the "sex" column to binary values (1 for male, 0 for female).
+    6. Fills any remaining missing values in the "age", "sex", and computed creatinine columns with 0.
+    7. Separates the target variable "aki" (if present) from the features.
+    Args:
+        data (str): The file path to the CSV data file.
+    Returns:
+        tuple: A tuple containing:
+            - X (pd.DataFrame): The preprocessed feature data.
+            - y (pd.Series or None): The target variable if "aki" is present in the data, otherwise None.
+    """
+    data = pd.read_csv(data)
+
+    # Extract the creatinine results, and all columns that contain the word "creatinine"
     creatinine_results = [col for col in data.columns if "creatinine_result" in col]
     all_creatinine_cols = [col for col in data.columns if "creatinine" in col]
     
+    # Compute the mean, min, max, and variance of the creatinine results
     data["creatinine_mean"] = data[creatinine_results].mean(axis=1)
     data["creatinine_min"] = data[creatinine_results].min(axis=1)
     data["creatinine_max"] = data[creatinine_results].max(axis=1)
     data["creatinine_variance"] = data[creatinine_results].var(axis=1)
-    data["creatinine_median"] = data[creatinine_results].median(axis=1)
 
+    # Once computed, drop the original creatinine columns to 
+    # keep only the computed features
     data = data.drop(columns=all_creatinine_cols)
 
+    # Fill missing values in the "age" column with the median age
     data["age"] = data["age"].fillna(data["age"].median())
+
+    # Encode the 'sex' column as binary values
     data["sex"] = data["sex"].map({"m": 1, "f": 0})
 
+    # Fill any remaining missing values with 0
     data["creatinine_variance"] = data["creatinine_variance"].fillna(0)
     data["creatinine_mean"] = data["creatinine_mean"].fillna(0)
     data["creatinine_min"] = data["creatinine_min"].fillna(0)
     data["creatinine_max"] = data["creatinine_max"].fillna(0)
-    data["creatinine_median"] = data["creatinine_median"].fillna(0)
-    data["age"] = data["age"].fillna(0)
-    data["sex"] = data["sex"].fillna(0)
 
+    # Separate the target variable if present. Return the 
+    # features and target variable, if present
     y = None
     X = data
     if "aki" in data.columns:
@@ -43,31 +66,42 @@ def preprocess_data(input_file):
     return X, y
 
 def main():
+    """
+    Trains a logistic regression model to predict the presence of AKI.
+    Logistic regression is a powerful but efficient linear model that is often used as a 
+    baseline model for binary classification tasks, hance making it a suitable choice here.
+    """
     parser = argparse.ArgumentParser()
     parser.add_argument("--input", default="test.csv")
     parser.add_argument("--output", default="aki.csv")
     parser.add_argument("--train", default="data/training.csv")
     flags = parser.parse_args()
 
+
+    # Preprocess the training data
     X_train, y_train = preprocess_data(flags.train)
     X_test, y_test = preprocess_data(flags.input)
 
+    # Standardize the data
     scaler = StandardScaler()
     X_train = scaler.fit_transform(X_train)
     X_test = scaler.transform(X_test)
     
+    # Train a logistic regression model with balanced class weights. 
     model = LogisticRegression(class_weight="balanced", random_state=42)
     model.fit(X_train, y_train)
 
     y_pred = model.predict(X_test)
-    preds = ["y" if p == 1 else "n" for p in y_pred]
 
+    # Make the predictions and write to a CSV file
+    preds = ["y" if p == 1 else "n" for p in y_pred]
     with open(flags.output, "w", newline="") as f:
         writer = csv.writer(f)
         writer.writerow(["aki"])
         for pred in preds:
             writer.writerow([pred])
 
+    # Evalute the model by computing the F3 score
     if y_test is not None:
         f3 = fbeta_score(y_test, y_pred, beta=3)
         print("F3 score: ", f3)