房价预测《基础版，测试》

#coding=utf8

import numpy as np
import pandas as pd
from sklearn.linear_model import Ridge
from sklearn.model_selection import cross_val_score
import matplotlib.pyplot as plt
from sklearn.ensemble import RandomForestRegressor

#不要第一列id，只是作为索引
train_df = pd.read_csv('./input/train.csv', index_col=0)
test_df = pd.read_csv('./input/test.csv', index_col=0)
#label本身并不平滑。为了我们分类器的学习更加准确，我们会首先把label给“平滑化”（正态化），如果miss掉，导致自己的结果总是达不到一定标准。这里我们使用最有逼格的log1p, 也就是 log(x+1)，避免了复值的问题。如果我们这里把数据都给平滑化了，那么最后算结果的时候，要记得把预测到的平滑数据给变回去。按照“怎么来的怎么去”原则，log1p()就需要expm1(); 同理，log()就需要exp(), ... etc.

prices = pd.DataFrame({'price':train_df['SalePrice'], 'log(price + 1)':np.log1p(train_df['SalePrice'])})

#print train_df.columns
#prices.hist()
#print 'ok'
y_train = np.log1p(train_df.pop('SalePrice'))
#print y_train.shape
#print train_df.index
all_df = pd.concat((train_df,test_df), axis=0)
#变量转换

#print all_df['MSSubClass'].dtypes
all_df['MSSubClass'] = all_df['MSSubClass'].astype(str)
#print all_df.shape
#print all_df['MSSubClass'].value_counts()
#print all_df['MSSubClass'].dtypes
#print pd.get_dummies(all_df['MSSubClass'], prefix='MSSubClass').head()
#当我们用numerical来表达categorical的时候，要注意，数字本身有大小的含义，所以乱用数字会给之后的模型学习带来麻烦。于是我们可以用One-Hot的方法来表达category。
#pandas自带的get_dummies方法，一键做到One-Hot。
#把所有的category数据，都给One-Hot了
all_dummy_df = pd.get_dummies(all_df)
#print all_dummy_df.head()
#print all_dummy_df.isnull().sum().sort_values(ascending=False).head(10)
#处理缺失值
mean_cols = all_dummy_df.mean()
#print mean_cols
all_dummy_df = all_dummy_df.fillna(mean_cols)
#print all_dummy_df.isnull().sum().sum()
#标准化numerical数据,这里，我们当然不需要把One-Hot的那些0/1数据给标准化。我们的目标应该是那些本来就是numerical的数据：
#先来看看 哪些是numerical的
numeric_cols = all_df.columns[all_df.dtypes != 'object']
#print numeric_cols
#print train_df.index
numeric_col_means = all_dummy_df.loc[:, numeric_cols].mean()
numeric_col_std = all_dummy_df.loc[:, numeric_cols].std()
all_dummy_df.loc[:, numeric_cols] = (all_dummy_df.loc[:, numeric_cols] - numeric_col_means) / numeric_col_std

dummy_train_df = all_dummy_df.loc[train_df.index]
dummy_test_df = all_dummy_df.loc[test_df.index]
#print train_df.index
#print test_df.index
#print dummy_train_df.shape
#print dummy_test_df.shape
#print type(dummy_train_df)

X_train = dummy_train_df.values
X_test = dummy_test_df.values
#print type(X_train)

print X_train.shape
alphas = np.logspace(-3, 2, 50)
test_scores = []
for alpha in alphas:
    clf = Ridge(alpha)
    test_score = np.sqrt(-cross_val_score(clf, X_train, y_train, cv=10, scoring='neg_mean_squared_error'))
    test_scores.append(np.mean(test_score))

plt.plot(alphas, test_scores)
plt.title('Alpha vs CV Error')

max_features = [.1, .3, .5, .7, .9, .99]
test_scores = []
for max_feat in max_features:
    clf = RandomForestRegressor(n_estimators=200, max_features=max_feat)
    test_score = np.sqrt(-cross_val_score(clf, X_train, y_train, cv=5, scoring='neg_mean_squared_error'))
    test_scores.append(np.mean(test_score))

plt.plot(max_features, test_scores)
plt.title("Max Features vs CV Error")

#Ensemble
ridge = Ridge(alpha=15)
rf = RandomForestRegressor(n_estimators=500, max_features=.3)

ridge.fit(X_train, y_train)
rf.fit(X_train, y_train)

y_ridge = np.expm1(ridge.predict(X_test))
y_rf = np.expm1(rf.predict(X_test))
y_final = (y_ridge + y_rf) / 2