本文整理汇总了Python中sklearn.ensemble.RandomForestRegressor.warm_start方法的典型用法代码示例。如果您正苦于以下问题：Python RandomForestRegressor.warm_start方法的具体用法？Python RandomForestRegressor.warm_start怎么用？Python RandomForestRegressor.warm_start使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类sklearn.ensemble.RandomForestRegressor的用法示例。
在下文中一共展示了RandomForestRegressor.warm_start方法的1个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: train_and_predict
# 需要导入模块: from sklearn.ensemble import RandomForestRegressor [as 别名]
# 或者: from sklearn.ensemble.RandomForestRegressor import warm_start [as 别名]
def train_and_predict(train_data_file, test_data_file, target_col, test_pred_file,
test_data_file2, test_pred_file2,
model_type, model_file, fit_args, test_metric, na_fill_value,
silent, skip_mapping, load_model, train_filter, metric_type, load_type,
bootstrap, bootstrap_seed, weight_col):
start = timeit.default_timer()
train_x = load_pd_df(train_data_file)
len_train_before = len(train_x)
train_x = data_filter(train_x, train_filter)
if not silent:
print "Train has %d instances (was %d before filtering)" % (len(train_x), len_train_before)
mappings = None if skip_mapping else dict()
if mappings is not None:
data_all = train_x.append(load_pd_df(test_data_file))
if test_data_file2 is not None:
data_all = data_all.append(load_pd_df(test_data_file2))
if not silent:
print "Mapping unkown and category values..."
for col in train_x.columns:
if col not in ['target_col']:
if data_all[col].dtype == np.dtype('object'):
s = np.unique(data_all[col].fillna(na_fill_value).values)
mappings[col] = pd.Series([x[0] for x in enumerate(s)], index=s)
train_x[col] = train_x[col].map(mappings[col]).fillna(na_fill_value)
else:
train_x[col] = train_x[col].fillna(na_fill_value)
del data_all
train_y = train_x[target_col]
del train_x[target_col]
extra_fit_args = dict()
if weight_col is not None:
extra_fit_args['sample_weight'] = train_x[weight_col].values
del train_x[weight_col]
if 0 < bootstrap < 1.0:
if bootstrap_seed is not None:
if not silent:
print "Setting bootstrap seed to %d" % bootstrap_seed
np.random.seed(bootstrap_seed)
random.seed(bootstrap_seed)
bootstrap_len = int(math.floor(bootstrap * len(train_x)))
bootstrap_ix = random.sample(range(len(train_x)), bootstrap_len)
train_x = train_x.iloc[bootstrap_ix]
train_x.reset_index()
train_y = train_y.iloc[bootstrap_ix]
train_y.reset_index()
x_cols = train_x.columns
feat_importance_fun = lambda (fitted_model): fitted_model.feature_importances_
predict = lambda (fitted_model, pred_x): fitted_model.predict(pred_x)
staged_predict = lambda (fitted_model, pred_x): [predict((fitted_model, pred_x))]
model = None
if load_model and os.path.exists(model_file):
if not silent:
print "Loading model %s" % model_file
model = load_model_bin(model_file=model_file)
if model_type == "RandomForestRegressor":
if model is None:
model = RandomForestRegressor(**fit_args)
model.fit(X=train_x, y=train_y, **extra_fit_args)
predict = lambda (fitted_model, pred_x): continuous_predict(model=fitted_model, x=pred_x)
elif model_type == "RandomForestClassifier":
if model is None:
model = RandomForestClassifier(**fit_args)
model.fit(X=train_x, y=train_y, **extra_fit_args)
predict = lambda (fitted_model, pred_x): pred_proba(model=fitted_model, x=pred_x)
staged_predict = lambda (fitted_model, pred_x): [predict((fitted_model, pred_x))]
elif model_type == "ExtraTreesRegressor":
if model is None:
model = ExtraTreesRegressor(**fit_args)
model.fit(X=train_x, y=train_y, **extra_fit_args)
predict = lambda (fitted_model, pred_x): continuous_predict(model=fitted_model, x=pred_x)
elif model_type == "ExtraTreesClassifier":
if model is None:
model = ExtraTreesClassifier(**fit_args)
model.fit(X=train_x, y=train_y, **extra_fit_args)
predict = lambda (fitted_model, pred_x): pred_proba(model=fitted_model, x=pred_x)
staged_predict = lambda (fitted_model, pred_x): [predict((fitted_model, pred_x))]
elif model_type == "GradientBoostingRegressor":
if model is None:
model = GradientBoostingRegressor(**fit_args)
model.fit(X=train_x, y=train_y, **extra_fit_args)
elif load_type == "fit_more":
model.warm_start = True
model.n_estimators += fit_args['n_estimators']
model.fit(X=train_x, y=train_y)
predict = lambda (fitted_model, pred_x): continuous_predict(model=fitted_model, x=pred_x)
staged_predict = lambda (fitted_model, pred_x): staged_pred_continuous(model=fitted_model, x=pred_x)
if load_type == "pred_at" and fit_args['n_estimators'] < model.n_estimators:
if not silent:
#.........这里部分代码省略.........