Commit Live PR

q01_myXGBoost/build.py

@@ -1,8 +1,11 @@
+# %load q01_myXGBoost/build.py
 import pandas as pd
 from xgboost import XGBClassifier
 from sklearn.model_selection import train_test_split
 from sklearn.model_selection import GridSearchCV
 from sklearn.metrics import accuracy_score
+from unittest import TestCase
+from inspect import getargspec
 
 # load data
 dataset = pd.read_csv('data/loan_clean_data.csv')
@@ -11,13 +14,38 @@
 y = dataset.iloc[:, -1]
 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=9)
 
-param_grid1 = {"max_depth": [2, 3, 4, 5, 6, 7, 9, 11],
-               "min_child_weight": [4, 6, 7, 8],
-               "subsample": [0.6, .7, .8, .9, 1],
-               "colsample_bytree": [0.6, .7, .8, .9, 1]
+param_grid1 = {'max_depth': [2, 3, 4, 5, 6, 7, 9, 11],
+               'min_child_weight': [4, 6, 7, 8],
+               'subsample': [0.6, .7, .8, .9, 1],
+               'colsample_bytree': [0.6, .7, .8, .9, 1]
                }
 
 
 # Write your solution here :
+def myXGBoost(X_train, X_test, y_train, y_test, model, param_grid, Kfold=3, **kwargs):
+    gs_cv = GridSearchCV(estimator = model, param_grid= param_grid, cv = Kfold)
+    gs_cv.fit(X_train, y_train)
+
+    # make predictions for test data
+    y_pred = gs_cv.predict(X_test)
+    predictions = [round(value) for value in y_pred]
 
+    accuracy = accuracy_score(y_test, predictions)
+    #print('Accuracy: %.2f%%' % (accuracy * 100.0))
+
+    return accuracy, gs_cv.best_params_
+
+xgb = XGBClassifier(seed=9)
+gs_cv_accuracy, gs_cv_best_params = myXGBoost(X_train, X_test, y_train, y_test, xgb, param_grid1, 3)
+expected_best_params = {'subsample': 0.8, 'colsample_bytree': 0.7, 'max_depth': 2, 'min_child_weight': 4}
+expected_accuracy = 0.796703296703
+
+print (gs_cv_accuracy)
+print (gs_cv_best_params)
+
+args = getargspec(myXGBoost)
+print (len(args[0]))
+print (args[3])
+print (type(gs_cv_accuracy))
+print (type(gs_cv_best_params))
 

q02_param2/build.py

@@ -17,4 +17,7 @@
                }
 
 
-# Write your solution here :
+def param2(X_train, X_test, y_train, y_test, xgb, param_grid, **kwargs):
+    #Include parameters used for earlier call as well.
+    accuracy, best_params_ = myXGBoost(X_train, X_test, y_train, y_test, xgb, param_grid, colsample_bytree=0.7, subsample=0.8, max_depth=2, min_child_weight=4)
+    return accuracy, best_params_

q03_xgboost/build.py

@@ -12,6 +12,16 @@
 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=9)
 
 
-# Write your solution here :
-
-
+def xgboost(X_train, X_test, y_train, y_test, **kwargs):
+
+    xgb = XGBClassifier(subsample=0.8,
+      colsample_bytree=0.7, max_depth=2,
+      min_child_weight=4, reg_alpha=0, reg_lambda=1.0,
+     gamma=0,n_estimators=100,learning_rate=0.01)
+    xgb.fit(X_train, y_train)
+    y_pred = xgb.predict(X_test)
+    y_pred = predictions = [round(value) for value in y_pred]
+
+    acc = accuracy_score(y_test, y_pred)
+
+    return acc