def buildForest(self, X_train, y_train):
NUM_TREES = 100
NUM_JOBS = 1
FEATURES_IN_EACH_TREE = "sqrt"
rf = RandomForestRegressor(n_estimators=NUM_TREES, verbose=1, n_jobs=NUM_JOBS, max_features=FEATURES_IN_EACH_TREE, oob_score=True, max_depth=25)
rf.fit_transform(X_train, y_train)
return rf
def predict_on_test(): sample = pd.read_csv(join(SAMPLES_FILE_PATH, "sample_train.csv")) test = pd.read_csv(join(SAMPLES_FILE_PATH, "sample_test.csv")) preprocessed = Preprocess(sample) rf = RandomForestRegressor(n_estimators = 100, criterion = "mse", bootstrap = True, max_features = 'sqrt', depth = 40) rf.fit_transform(X = preprocessed.features, y = preprocessed.labels.values.ravel()) test_preprocessed = Preprocess(test) predicted_values = rf.predict(test_preprocessed.features) error_rate, _ = benchmark(predicted_values.ravel(), test_preprocessed.labels.values) print "Mean Square Prediction Erorr = %s" % MSE_prediction(predicted_values, test_preprocessed.labels) plot_feature_importances(preprocessed.features.columns.values, rf.feature_importances_)
def cross_validate_depth():
sample = pd.read_csv(join(SAMPLES_FILE_PATH, "sample_train.csv"))
preprocessed = Preprocess(sample)
depths = (2, 40, 60, 80)
oob_scores = []
for depth in depths:
rf = RandomForestRegressor(n_estimators = 60, criterion = "mse", bootstrap = True, oob_score = True, max_features = 'sqrt', max_depth = depth)
rf.fit_transform(X = preprocessed.features, y = preprocessed.labels.values.ravel())
score = 1.0 - rf.oob_score_
oob_scores.append(score)
print "Out-of-Bag Error for Depth %s: %s" % (depth, score)
pdb.set_trace()
plot_oob_error_depth(depths, oob_scores)
dftrain_X, feature_names = dense2sparse(dftrain_Xdense)
dftest_X, feature_names = dense2sparse(dftest_Xdense)
# X is your feature space, Y is your target variable
dftrain_y = df1[df1.date <= date1]['FanDuelPts']
dftest_y = df1[df1.date > date1]['FanDuelPts']
#==============================================================================
# Predict last two weeks
#==============================================================================
from sklearn.ensemble import RandomForestRegressor
rf1 = RandomForestRegressor(verbose=True) # Parameters need to be tuned
rf1.fit_transform(dftrain_X,dftrain_y) # Train the model
rf1_preds = rf1.predict(dftest_X) # Predict against the test set
# Performance metrics
from sklearn.metrics import mean_absolute_error, mean_squared_error
mean_absolute_error(dftest_y,rf1_preds)
np.sqrt(mean_squared_error(dftest_y,rf1_preds))
# Plot model results
dfplot = pd.merge(dftest_y.to_frame('Actual'), pd.DataFrame(rf1_preds,
columns=['Pred']), left_index=True, right_index=True)
dfplot.plot()
import matplotlib.pyplot as plt
plt.plot(np.arange(1,len(rf1_preds),),rf1_preds)
import matplotlib.pyplot as plt
import pandas as pd
# Importing the dataset
dataset = pd.read_csv('Position_Salaries.csv')
X = dataset.iloc[:, [1]].values
y = dataset.iloc[:,[2]].values
# Splitting the dataset into the Training set and Test set
"""from sklearn.cross_validation import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)"""
# Feature Scaling
"""from sklearn.preprocessing import StandardScaler
sc_X = StandardScaler()
X_train = sc_X.fit_transform(X_train)
X_test = sc_X.transform(X_test)"""
from sklearn.ensemble import RandomForestRegressor
rfr= RandomForestRegressor(n_estimators=50, random_state=0)
rfr.fit_transform(X,y)
y_pred= rfr.predict(6.5)
y_pred
X_grid=np.arange(min(X),max(X),0.01)
X_grid=X_grid.reshape((len(X_grid),1))
plt.scatter(X,y, color='blue')
plt.plot(X_grid, rfr.predict(X_grid), color='green')
plt.title('RFR')
plt.xlabel('positionsalaries')
plt.ylabel('salary')
plt.show()
# parameters
n_estimators = 200
max_depth = 25
min_samples_split = 15
min_samples_leaf = 2
# Random forest classifier
clf = RandomForestRegressor(n_estimators=n_estimators,
max_depth=max_depth,
min_samples_split=min_samples_split,
min_samples_leaf=min_samples_leaf)
# It is trained of 2 Epochs
X = np.concatenate((X, X), axis=0)
Y = np.concatenate((Y, Y), axis=0)
clf.fit_transform(X, Y)
print "Feature Importance ranking", clf.feature_importances_
# Saving the model for predictions
file_name = "model-final"
print "---- Saved at -----" + file_name
fileObject = open(file_name, 'wb')
#
pickle.dump(clf, fileObject)
fileObject.close()
print("--- %s seconds ---" % (time.time() - start_time))