def train_lr_model(train_file, model_coef, model_file, feature_num_file):
"""
Args:
train_file: process file for lr train
model_coef: w1 w2...
model_file:model pkl
feature_num_file: file to record num of feature
"""
total_feature_num = get_feature_num(feature_num_file)
train_label = np.genfromtxt(train_file,
dtype=np.int32,
delimiter=",",
usecols=-1)
feature_list = range(total_feature_num)
train_feature = np.genfromtxt(train_file,
dtype=np.int32,
delimiter=",",
usecols=feature_list)
lr_cf = lrcv(Cs=[1],
penalty="l2",
tol=0.0001,
max_iter=500,
cv=5,
scoring="roc_auc").fit(train_feature, train_label)
scores = lr_cf.scores_[1]
print(scores)
print("diff:%s" % (",".join([str(ele) for ele in scores.mean(axis=0)])))
print("Accuracy:%s (+-%0.2f)" % (scores.mean(), scores.std() * 2))
coef = lr_cf.coef_[0]
fw = open(model_coef, "w+")
fw.write(",".join(str(ele) for ele in coef))
fw.close()
joblib.dump(lr_cf, model_file)
'verbose': 0
}
gbm = lgb.train(params,
train,
num_boost_round=100,
valid_sets=test,
early_stopping_rounds=10)
gbm.save_model('gbm.txt')
y_pred = gbm.predict(test_x, num_iteration=gbm.best_iteration)
print(y_pred)
# print('The roc of prediction is:', roc_auc_score(test_y, y_pred))
y_pred = gbm.predict(feature, num_iteration=gbm.best_iteration)
feature['leaf'] = y_pred
lr_cf = lrcv(Cs=[1], penalty="l2", tol=0.0001, max_iter=200,
cv=5).fit(feature.as_matrix(),
np.array(label.values.tolist()).reshape((-1, 1)))
item = pd.read_csv('data/tianchi_fresh_comp_train_user.csv')
item['time'] = item['time'].apply(lambda x: timestamp(x))
item = item[['item_id',
'item_category']].drop_duplicates('item_id').astype('int32')
cfp = pd.read_csv('cf_predict.csv')
cfp = cfp.merge(item, on='item_id', how='left')
cfp = cfp[['user_id', 'item_id', 'item_category']]
cfp['time'] = int(time.time())
feature = cfp[['user_id', 'item_id', 'item_category', 'time']]
y_pred = gbm.predict(feature, num_iteration=gbm.best_iteration, pred_leaf=True)
y_pred = y_pred.tolist()
print('leaf %s' % y_pred)