feat: add more templates for kaggle (#291)

* init for forest-cover-type-prediction

* add nn model for forest-cover-type-prediction

* add cross_validation for forest-cover-type-prediction

* edit path to file

* CI issues

* CI Issue

* edit dir name

* fix a bug in s4e8 ensemble & init spaceship-titanic

* add nn model for s4e8 & spaceship-titanic

* init for s4e9

* ci issues

* ci issue

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/cross_validation.py

@@ -0,0 +1,101 @@
+import importlib.util
+import random
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+from scipy import stats
+from sklearn.impute import SimpleImputer
+from sklearn.metrics import accuracy_score
+from sklearn.model_selection import KFold
+
+# Set random seed for reproducibility
+SEED = 42
+random.seed(SEED)
+np.random.seed(SEED)
+DIRNAME = Path(__file__).absolute().resolve().parent
+
+
+def import_module_from_path(module_name, module_path):
+    spec = importlib.util.spec_from_file_location(module_name, module_path)
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+# 1) Preprocess the data
+data_df = pd.read_csv(
+    "/data/userdata/v-haoranpan/RD-Agent/git_ignore_folder/data/forest-cover-type-prediction/train.csv"
+)
+data_df = data_df.drop(["Id"], axis=1)
+
+X_train = data_df.drop(["Cover_Type"], axis=1)
+y_train = data_df["Cover_Type"] - 1
+
+# Set up KFold
+kf = KFold(n_splits=5, shuffle=True, random_state=SEED)
+
+# Store results
+accuracies = []
+
+# 3) Train and evaluate using KFold
+fold_number = 1
+for train_index, valid_index in kf.split(X_train):
+    print(f"Starting fold {fold_number}...")
+
+    X_train_l, X_valid_l = [], []  # Reset feature lists for each fold
+    X_tr, X_val = X_train.iloc[train_index], X_train.iloc[valid_index]
+    y_tr, y_val = y_train.iloc[train_index], y_train.iloc[valid_index]
+
+    # Feature engineering
+    for f in DIRNAME.glob("feature/feat*.py"):
+        cls = import_module_from_path(f.stem, f).feature_engineering_cls()
+        cls.fit(X_tr)
+        X_train_f = cls.transform(X_tr)
+        X_valid_f = cls.transform(X_val)
+
+        X_train_l.append(X_train_f)
+        X_valid_l.append(X_valid_f)
+
+    X_tr = pd.concat(X_train_l, axis=1, keys=[f"feature_{i}" for i in range(len(X_train_l))])
+    X_val = pd.concat(X_valid_l, axis=1, keys=[f"feature_{i}" for i in range(len(X_valid_l))])
+
+    print("Shape of X_tr: ", X_tr.shape, " Shape of X_val: ", X_val.shape)
+
+    # Replace inf and -inf with NaN
+    X_tr.replace([np.inf, -np.inf], np.nan, inplace=True)
+    X_val.replace([np.inf, -np.inf], np.nan, inplace=True)
+
+    # Impute missing values
+    imputer = SimpleImputer(strategy="mean")
+    X_tr = pd.DataFrame(imputer.fit_transform(X_tr), columns=X_tr.columns)
+    X_val = pd.DataFrame(imputer.transform(X_val), columns=X_val.columns)
+
+    # Remove duplicate columns
+    X_tr = X_tr.loc[:, ~X_tr.columns.duplicated()]
+    X_val = X_val.loc[:, ~X_val.columns.duplicated()]
+
+    # Train the model
+    model_l = []  # list[tuple[model, predict_func]]
+    for f in DIRNAME.glob("model/model*.py"):
+        m = import_module_from_path(f.stem, f)
+        model_l.append((m.fit(X_tr, y_tr, X_val, y_val), m.predict))
+
+    # Evaluate the model on the validation set
+    y_valid_pred_l = []
+    for model, predict_func in model_l:
+        y_valid_pred = predict_func(model, X_val)
+        y_valid_pred_l.append(y_valid_pred)
+
+    # Majority vote ensemble
+    y_valid_pred_ensemble = stats.mode(y_valid_pred_l, axis=0)[0].flatten()
+
+    # Compute metrics
+    accuracy = accuracy_score(y_val, y_valid_pred_ensemble)
+    accuracies.append(accuracy)
+    print(f"Fold {fold_number} accuracy: {accuracy}")
+
+    fold_number += 1
+
+# Print average accuracy
+print(f"Average accuracy across folds: {np.mean(accuracies)}")

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/fea_share_preprocess.py

@@ -0,0 +1,72 @@
+import os
+
+import numpy as np
+import pandas as pd
+from sklearn.impute import SimpleImputer
+from sklearn.model_selection import train_test_split
+
+
+def prepreprocess():
+    """
+    This method loads the data, drops the unnecessary columns, and splits it into train and validation sets.
+    """
+    # Load and preprocess the data
+    data_df = pd.read_csv("/kaggle/input/train.csv")
+    data_df = data_df.drop(["Id"], axis=1)
+
+    X = data_df.drop(["Cover_Type"], axis=1)
+    y = data_df["Cover_Type"] - 1
+
+    # Split the data into training and validation sets
+    X_train, X_valid, y_train, y_valid = train_test_split(X, y, test_size=0.20, random_state=42)
+
+    return X_train, X_valid, y_train, y_valid
+
+
+def preprocess_script():
+    """
+    This method applies the preprocessing steps to the training, validation, and test datasets.
+    """
+    if os.path.exists("X_train.pkl"):
+        X_train = pd.read_pickle("X_train.pkl")
+        X_valid = pd.read_pickle("X_valid.pkl")
+        y_train = pd.read_pickle("y_train.pkl")
+        y_valid = pd.read_pickle("y_valid.pkl")
+        X_test = pd.read_pickle("X_test.pkl")
+        ids = pd.read_pickle("ids.pkl")
+
+        return X_train, X_valid, y_train, y_valid, X_test, ids
+
+    X_train, X_valid, y_train, y_valid = prepreprocess()
+
+    # Load and preprocess the test data
+    submission_df = pd.read_csv("/kaggle/input/test.csv")
+    ids = submission_df["Id"]
+    X_test = submission_df.drop(["Id"], axis=1)
+
+    return X_train, X_valid, y_train, y_valid, X_test, ids
+
+
+def clean_and_impute_data(X_train, X_valid, X_test):
+    """
+    Handles inf and -inf values by replacing them with NaN,
+    then imputes missing values using the mean strategy.
+    Also removes duplicate columns.
+    """
+    # Replace inf and -inf with NaN
+    X_train.replace([np.inf, -np.inf], np.nan, inplace=True)
+    X_valid.replace([np.inf, -np.inf], np.nan, inplace=True)
+    X_test.replace([np.inf, -np.inf], np.nan, inplace=True)
+
+    # Impute missing values
+    imputer = SimpleImputer(strategy="mean")
+    X_train = pd.DataFrame(imputer.fit_transform(X_train), columns=X_train.columns)
+    X_valid = pd.DataFrame(imputer.transform(X_valid), columns=X_valid.columns)
+    X_test = pd.DataFrame(imputer.transform(X_test), columns=X_test.columns)
+
+    # Remove duplicate columns
+    X_train = X_train.loc[:, ~X_train.columns.duplicated()]
+    X_valid = X_valid.loc[:, ~X_valid.columns.duplicated()]
+    X_test = X_test.loc[:, ~X_test.columns.duplicated()]
+
+    return X_train, X_valid, X_test

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/feature/feature.py

@@ -0,0 +1,23 @@
+import pandas as pd
+
+"""
+Here is the feature engineering code for each task, with a class that has a fit and transform method.
+Remember
+"""
+
+
+class IdentityFeature:
+    def fit(self, train_df: pd.DataFrame):
+        """
+        Fit the feature engineering model to the training data.
+        """
+        pass
+
+    def transform(self, X: pd.DataFrame):
+        """
+        Transform the input data.
+        """
+        return X
+
+
+feature_engineering_cls = IdentityFeature

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/model/model_nn.py

@@ -0,0 +1,78 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.utils.data import DataLoader, TensorDataset
+from tqdm import tqdm
+
+# Check if a GPU is available
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+
+# Modified model for multi-class classification
+class HybridFeatureInteractionModel(nn.Module):
+    def __init__(self, num_features, num_classes):
+        super(HybridFeatureInteractionModel, self).__init__()
+        self.fc1 = nn.Linear(num_features, 128)
+        self.bn1 = nn.BatchNorm1d(128)
+        self.fc2 = nn.Linear(128, 64)
+        self.bn2 = nn.BatchNorm1d(64)
+        self.fc3 = nn.Linear(64, num_classes)  # Output nodes equal to num_classes
+        self.dropout = nn.Dropout(0.3)
+
+    def forward(self, x):
+        x = F.relu(self.bn1(self.fc1(x)))
+        x = F.relu(self.bn2(self.fc2(x)))
+        x = self.dropout(x)
+        x = self.fc3(x)  # No activation here, use CrossEntropyLoss
+        return x
+
+
+# Training function
+def fit(X_train, y_train, X_valid, y_valid):
+    num_features = X_train.shape[1]
+    num_classes = len(np.unique(y_train))  # Determine number of classes
+    model = HybridFeatureInteractionModel(num_features, num_classes).to(device)
+    criterion = nn.CrossEntropyLoss()  # Use CrossEntropyLoss for multi-class
+    optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
+
+    # Convert to TensorDataset and create DataLoader
+    train_dataset = TensorDataset(
+        torch.tensor(X_train.to_numpy(), dtype=torch.float32), torch.tensor(y_train.to_numpy(), dtype=torch.long)
+    )
+    valid_dataset = TensorDataset(
+        torch.tensor(X_valid.to_numpy(), dtype=torch.float32), torch.tensor(y_valid.to_numpy(), dtype=torch.long)
+    )
+    train_loader = DataLoader(train_dataset, batch_size=32, shuffle=True)
+    valid_loader = DataLoader(valid_dataset, batch_size=32, shuffle=False)
+
+    # Train the model
+    model.train()
+    for epoch in range(5):  # just for quick run
+        print(f"Epoch {epoch + 1}/5")
+        epoch_loss = 0
+        for X_batch, y_batch in tqdm(train_loader, desc="Training", leave=False):
+            X_batch, y_batch = X_batch.to(device), y_batch.to(device)
+            optimizer.zero_grad()
+            outputs = model(X_batch)
+            loss = criterion(outputs, y_batch)
+            loss.backward()
+            optimizer.step()
+            epoch_loss += loss.item()
+        print(f"End of epoch {epoch + 1}, Avg Loss: {epoch_loss / len(train_loader):.4f}")
+
+    return model
+
+
+# Prediction function
+def predict(model, X):
+    model.eval()
+    predictions = []
+    with torch.no_grad():
+        X_tensor = torch.tensor(X.values, dtype=torch.float32).to(device)
+        for i in tqdm(range(0, len(X_tensor), 32), desc="Predicting", leave=False):
+            batch = X_tensor[i : i + 32]
+            pred = model(batch)
+            pred = torch.argmax(pred, dim=1).cpu().numpy()  # Use argmax to get class
+            predictions.extend(pred)
+    return np.array(predictions)

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/model/model_randomforest.py

@@ -0,0 +1,53 @@
+"""
+Motivation of the model:
+The Random Forest model is chosen for its robustness and ability to handle large datasets with higher dimensionality.
+It reduces overfitting by averaging multiple decision trees and typically performs well out of the box, making it a good
+baseline model for many classification tasks.
+"""
+
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.metrics import accuracy_score
+
+
+def select(X: pd.DataFrame) -> pd.DataFrame:
+    """
+    Select relevant features. To be used in fit & predict function.
+    """
+    # For now, we assume all features are relevant. This can be expanded to feature selection logic.
+    return X
+
+
+def fit(X_train: pd.DataFrame, y_train: pd.Series, X_valid: pd.DataFrame, y_valid: pd.Series):
+    """
+    Define and train the Random Forest model. Merge feature selection into the pipeline.
+    """
+    # Initialize the Random Forest model
+    model = RandomForestClassifier(n_estimators=100, random_state=32, n_jobs=-1)
+
+    # Select features (if any feature selection is needed)
+    X_train_selected = select(X_train)
+    X_valid_selected = select(X_valid)
+
+    # Fit the model
+    model.fit(X_train_selected, y_train)
+
+    # Validate the model
+    y_valid_pred = model.predict(X_valid_selected)
+    accuracy = accuracy_score(y_valid, y_valid_pred)
+    print(f"Validation Accuracy: {accuracy:.4f}")
+
+    return model
+
+
+def predict(model, X):
+    """
+    Keep feature selection's consistency and make predictions.
+    """
+    # Select features (if any feature selection is needed)
+    X_selected = select(X)
+
+    # Predict using the trained model
+    y_pred = model.predict(X_selected)
+
+    return y_pred

rdagent/scenarios/kaggle/experiment/forest-cover-type-prediction_template/model/model_xgboost.py

@@ -0,0 +1,41 @@
+"""
+motivation  of the model
+"""
+
+import pandas as pd
+import xgboost as xgb
+
+
+def select(X: pd.DataFrame) -> pd.DataFrame:
+    # Ignore feature selection logic
+    return X
+
+
+def fit(X_train: pd.DataFrame, y_train: pd.DataFrame, X_valid: pd.DataFrame, y_valid: pd.DataFrame):
+    """Define and train the model. Merge feature_select"""
+    X_train = select(X_train)
+    X_valid = select(X_valid)
+    dtrain = xgb.DMatrix(X_train, label=y_train)
+    dvalid = xgb.DMatrix(X_valid, label=y_valid)
+
+    params = {
+        "objective": "multi:softmax",  # Use softmax for multi-class classification
+        "num_class": len(set(y_train)),  # Number of classes
+        "nthread": -1,
+    }
+    num_round = 20
+
+    evallist = [(dtrain, "train"), (dvalid, "eval")]
+    bst = xgb.train(params, dtrain, num_round, evallist)
+
+    return bst
+
+
+def predict(model, X):
+    """
+    Keep feature select's consistency.
+    """
+    X = select(X)
+    dtest = xgb.DMatrix(X)
+    y_pred = model.predict(dtest)
+    return y_pred.astype(int)