import pandas as pd
import numpy as np
from scipy import sparse
from sklearn.model_selection import train_test_split
import xgboost as xgb
from sklearn.model_selection import GridSearchCV
def optimal_n_rounds(xgb_model, xgb_matrix, max_n_estimators):
""" take the input model and xgb matrix (x and y values)
and determine the optimal number of trees via cross validation.
returns the number of trees """
cvresult = xgb.cv(xgb_model, x_values,
num_boost_round = max_n_estimators,
nfold = 5,
metrics = 'auc',
early_stopping_rounds = 50)
return cvresult.shape[0]
def optimal_params(xgb_model, x_vals, y_vals, xgb_param_grid):
""" take a model, predictor matrix and paramater grid and
return the optimal paramater set """
_gsearch = GridSearchCV(xgb_model, xgb_param_grid,
scoring='roc_auc',
n_jobs=4,
iid=False,
cv=3)
_gsearch.fit(x_vals, y_vals)
return _gsearch.best_params_
if __name__ == '__main__':
#load in the processed data from train_and_test_to_matrix.py
train_sparse = sparse.load_npz('sparse_train_punc.npz')
test_sparse = sparse.load_npz('sparse_test_punc.npz')
train = pd.read_csv('train.csv')
test = pd.read_csv('test.csv')
sub_file = pd.read_csv('sample_submission.csv')
to_predict = list(train.columns[2:])
for col in to_predict:
xgtrain_input = xgb.DMatrix(train_sparse, label=train[col].values)
xgb_initial = xgb.XGBClassifier(learning_rate =0.1,
n_estimators=1000,
max_depth=5,
min_child_weight=1,
gamma=0,
subsample=0.8,
colsample_bytree=0.8,
objective= 'binary:logistic',
scale_pos_weight=1)
opt_rounds = optimal_n_rounds(xgb_initial, xgtrain_input, 1000)
xgb_class_grid = xgb.XGBClassifier(n_estimators=opt_rounds,
gamma=0,
subsample=0.8,
objective= 'binary:logistic',
scale_pos_weight=1)
xgb_params = {'max_depth':[4, 6, 8],
'min_child_weight':[1, 4, 8],
'colsample_bytree': [0.8, 0.9], }
xgb_best = optimal_params(xgb_class_grid,
train_sparse,
train[col].values,
xgb_params)
xgb_final = xgb.XGBClassifier(gsearch_toxic.best_params_,
eta = 0.001,
n_estimators=5000,
gamma=0,
subsample=0.8,
objective= 'binary:logistic',
scale_pos_weight=1,
early_stopping_rounds = 50)
xgb_final.fit(train_sparse, train[col].values, eval_metric='auc')
optimal_predictions = xgb_final.predict(test_sparse)
sub_file[col] = optimal_predictions
sub_file.to_csv('cam_xgb_predictions.csv', index = False)
