3.洛杉矶房价预测的Baseline

• 代码地址：appke/Los-House-Prices: 洛杉矶房价预测

房价预测的Baseline

结构化数据的 回归问题

import pandas as pd
import numpy as np

train=pd.read_csv('datas/los_data.csv') # 训练集的读取

train.get_dtype_counts()

float64     3
int64      35
object     43
dtype: int64

train.head(3)

	Id	MSSubClass	MSZoning	LotFrontage	LotArea	Street	Alley	LotShape	LandContour	Utilities	...	PoolArea	PoolQC	Fence	MiscFeature	MiscVal	MoSold	YrSold	SaleType	SaleCondition	SalePrice
0	1	60	RL	65.0	8450	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	2	2008	WD	Normal	208500
1	2	20	RL	80.0	9600	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	5	2007	WD	Normal	181500
2	3	60	RL	68.0	11250	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	9	2008	WD	Normal	223500

3 rows × 81 columns

train.info() # 类型统计

train.describe() # 类型描述

	Id	MSSubClass	LotFrontage	LotArea	OverallQual	OverallCond	YearBuilt	YearRemodAdd	MasVnrArea	BsmtFinSF1	...	WoodDeckSF	OpenPorchSF	EnclosedPorch	3SsnPorch	ScreenPorch	PoolArea	MiscVal	MoSold	YrSold	SalePrice
count	1460.000000	1460.000000	1201.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1452.000000	1460.000000	...	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000	1460.000000
mean	730.500000	56.897260	70.049958	10516.828082	6.099315	5.575342	1971.267808	1984.865753	103.685262	443.639726	...	94.244521	46.660274	21.954110	3.409589	15.060959	2.758904	43.489041	6.321918	2007.815753	180921.195890
std	421.610009	42.300571	24.284752	9981.264932	1.382997	1.112799	30.202904	20.645407	181.066207	456.098091	...	125.338794	66.256028	61.119149	29.317331	55.757415	40.177307	496.123024	2.703626	1.328095	79442.502883
min	1.000000	20.000000	21.000000	1300.000000	1.000000	1.000000	1872.000000	1950.000000	0.000000	0.000000	...	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	1.000000	2006.000000	34900.000000
25%	365.750000	20.000000	59.000000	7553.500000	5.000000	5.000000	1954.000000	1967.000000	0.000000	0.000000	...	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	5.000000	2007.000000	129975.000000
50%	730.500000	50.000000	69.000000	9478.500000	6.000000	5.000000	1973.000000	1994.000000	0.000000	383.500000	...	0.000000	25.000000	0.000000	0.000000	0.000000	0.000000	0.000000	6.000000	2008.000000	163000.000000
75%	1095.250000	70.000000	80.000000	11601.500000	7.000000	6.000000	2000.000000	2004.000000	166.000000	712.250000	...	168.000000	68.000000	0.000000	0.000000	0.000000	0.000000	0.000000	8.000000	2009.000000	214000.000000
max	1460.000000	190.000000	313.000000	215245.000000	10.000000	9.000000	2010.000000	2010.000000	1600.000000	5644.000000	...	857.000000	547.000000	552.000000	508.000000	480.000000	738.000000	15500.000000	12.000000	2010.000000	755000.000000

8 rows × 38 columns

查看数据DataFrame的数据类型与空值的表现

如何统计数据的缺失率？

train.isnull().head()

	Id	MSSubClass	MSZoning	LotFrontage	LotArea	Street	Alley	LotShape	LandContour	Utilities	...	PoolArea	PoolQC	Fence	MiscFeature	MiscVal	MoSold	YrSold	SaleType	SaleCondition	SalePrice
0	False	False	False	False	False	False	True	False	False	False	...	False	True	True	True	False	False	False	False	False	False
1	False	False	False	False	False	False	True	False	False	False	...	False	True	True	True	False	False	False	False	False	False
2	False	False	False	False	False	False	True	False	False	False	...	False	True	True	True	False	False	False	False	False	False
3	False	False	False	False	False	False	True	False	False	False	...	False	True	True	True	False	False	False	False	False	False
4	False	False	False	False	False	False	True	False	False	False	...	False	True	True	True	False	False	False	False	False	False

5 rows × 81 columns

1. 那些列有缺失？
2. 缺失了多少？
3. 缺失率是多少
4. 可否安缺失率由大到小排序

nullSum = train.isnull().sum()
train.isnull().sum()

tmp = train.isnull().sum()
nullSum_data = tmp[tmp>0]
tmp[tmp>0].shape

(19,)

nullSum_data

LotFrontage      259
Alley           1369
MasVnrType         8
MasVnrArea         8
BsmtQual          37
BsmtCond          37
BsmtExposure      38
BsmtFinType1      37
BsmtFinType2      38
Electrical         1
FireplaceQu      690
GarageType        81
GarageYrBlt       81
GarageFinish      81
GarageQual        81
GarageCond        81
PoolQC          1453
Fence           1179
MiscFeature     1406
dtype: int64

[i for i in nullSum_data]

[259,
 1369,
 8,
 8,
 37,
 37,
 38,
 37,
 38,
 1,
 690,
 81,
 81,
 81,
 81,
 81,
 1453,
 1179,
 1406]

[i/len(train) for i in nullSum_data]

[0.1773972602739726,
 0.9376712328767123,
 0.005479452054794521,
 0.005479452054794521,
 0.025342465753424658,
 0.025342465753424658,
 0.026027397260273973,
 0.025342465753424658,
 0.026027397260273973,
 0.0006849315068493151,
 0.4726027397260274,
 0.05547945205479452,
 0.05547945205479452,
 0.05547945205479452,
 0.05547945205479452,
 0.05547945205479452,
 0.9952054794520548,
 0.8075342465753425,
 0.963013698630137]

tmp=train.isnull().sum().sort_values()
tmp[tmp>0]

Electrical         1
MasVnrType         8
MasVnrArea         8
BsmtQual          37
BsmtCond          37
BsmtFinType1      37
BsmtFinType2      38
BsmtExposure      38
GarageQual        81
GarageFinish      81
GarageYrBlt       81
GarageType        81
GarageCond        81
LotFrontage      259
FireplaceQu      690
Fence           1179
Alley           1369
MiscFeature     1406
PoolQC          1453
dtype: int64

对连续型数据进行均值 | 中位数填充

train1=train

train1.mean().head()
# train1.median().head()

Id               730.500000
MSSubClass        56.897260
LotFrontage       70.049958
LotArea        10516.828082
OverallQual        6.099315
dtype: float64

# fillna函数会找到对应列的均值或者是中位数，对于该列进行相应的填充
train1 = train1.fillna(train1.mean())

# 空值统计函数
def getNullCount(train):
    tmp = train.isnull().sum()
    print(tmp[tmp>0].shape)

getNullCount(train1)

(16,)

对于离散型数据进行填充

朴素办法：将NaN这种空类型变成“None”或者是“NA”的字段

train1.head()

	Id	MSSubClass	MSZoning	LotFrontage	LotArea	Street	Alley	LotShape	LandContour	Utilities	...	PoolArea	PoolQC	Fence	MiscFeature	MiscVal	MoSold	YrSold	SaleType	SaleCondition	SalePrice
0	1	60	RL	65.0	8450	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	2	2008	WD	Normal	208500
1	2	20	RL	80.0	9600	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	5	2007	WD	Normal	181500
2	3	60	RL	68.0	11250	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	9	2008	WD	Normal	223500
3	4	70	RL	60.0	9550	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	2	2006	WD	Abnorml	140000
4	5	60	RL	84.0	14260	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	12	2008	WD	Normal	250000

5 rows × 81 columns

train1=train1.fillna("None")

train1.head()

	Id	MSSubClass	MSZoning	LotFrontage	LotArea	Street	Alley	LotShape	LandContour	Utilities	...	PoolArea	PoolQC	Fence	MiscFeature	MiscVal	MoSold	YrSold	SaleType	SaleCondition	SalePrice
0	1	60	RL	65.0	8450	Pave	None	Reg	Lvl	AllPub	...	0	None	None	None	0	2	2008	WD	Normal	208500
1	2	20	RL	80.0	9600	Pave	None	Reg	Lvl	AllPub	...	0	None	None	None	0	5	2007	WD	Normal	181500
2	3	60	RL	68.0	11250	Pave	None	IR1	Lvl	AllPub	...	0	None	None	None	0	9	2008	WD	Normal	223500
3	4	70	RL	60.0	9550	Pave	None	IR1	Lvl	AllPub	...	0	None	None	None	0	2	2006	WD	Abnorml	140000
4	5	60	RL	84.0	14260	Pave	None	IR1	Lvl	AllPub	...	0	None	None	None	0	12	2008	WD	Normal	250000

5 rows × 81 columns

getNullCount(train1) # 填充完毕

(0,)

制作训练集

1.y标签

2.数据里面有Id？pandas伪造出来的

3.答案分出来了，还把答案放进去？？

y = train1['SalePrice']# 标签label
train1 = train1.drop(['Id', 'SalePrice'],axis=1)

train1.head()

	MSSubClass	MSZoning	LotFrontage	LotArea	Street	Alley	LotShape	LandContour	Utilities	LotConfig	...	ScreenPorch	PoolArea	PoolQC	Fence	MiscFeature	MiscVal	MoSold	YrSold	SaleType	SaleCondition
0	60	RL	65.0	8450	Pave	None	Reg	Lvl	AllPub	Inside	...	0	0	None	None	None	0	2	2008	WD	Normal
1	20	RL	80.0	9600	Pave	None	Reg	Lvl	AllPub	FR2	...	0	0	None	None	None	0	5	2007	WD	Normal
2	60	RL	68.0	11250	Pave	None	IR1	Lvl	AllPub	Inside	...	0	0	None	None	None	0	9	2008	WD	Normal
3	70	RL	60.0	9550	Pave	None	IR1	Lvl	AllPub	Corner	...	0	0	None	None	None	0	2	2006	WD	Abnorml
4	60	RL	84.0	14260	Pave	None	IR1	Lvl	AllPub	FR2	...	0	0	None	None	None	0	12	2008	WD	Normal

5 rows × 79 columns

X = pd.get_dummies(train1) # 结构化数据

X.head()

	MSSubClass	LotFrontage	LotArea	OverallQual	OverallCond	YearBuilt	YearRemodAdd	MasVnrArea	BsmtFinSF1	BsmtFinSF2	...	SaleType_ConLw	SaleType_New	SaleType_Oth	SaleType_WD	SaleCondition_Abnorml	SaleCondition_AdjLand	SaleCondition_Alloca	SaleCondition_Family	SaleCondition_Normal	SaleCondition_Partial
0	60	65.0	8450	7	5	2003	2003	196.0	706	0	...	0	0	0	1	0	0	0	0	1	0
1	20	80.0	9600	6	8	1976	1976	0.0	978	0	...	0	0	0	1	0	0	0	0	1	0
2	60	68.0	11250	7	5	2001	2002	162.0	486	0	...	0	0	0	1	0	0	0	0	1	0
3	70	60.0	9550	7	5	1915	1970	0.0	216	0	...	0	0	0	1	1	0	0	0	0	0
4	60	84.0	14260	8	5	2000	2000	350.0	655	0	...	0	0	0	1	0	0	0	0	1	0

5 rows × 303 columns

X.shape

(1460, 303)

数据切分

分一些数据做评测

from sklearn.model_selection import train_test_split

X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.2,random_state=123)

X_train.shape

(1168, 303)

X_test.shape

(292, 303)

BaseLine模型的训练

import xgboost as xgb

xg_reg=xgb.XGBRegressor(objective='reg:linear',colsample_bytree=0.6,learning_rate=0.01,max_depth=8,alpha=10,n_estimators=700,sum_sample=0.7)

xg_reg.fit(X_train, y_train)

XGBRegressor(alpha=10, base_score=0.5, booster='gbtree', colsample_bylevel=1,
       colsample_bytree=0.6, gamma=0, importance_type='gain',
       learning_rate=0.01, max_delta_step=0, max_depth=8,
       min_child_weight=1, missing=None, n_estimators=700, n_jobs=1,
       nthread=None, objective='reg:linear', random_state=0, reg_alpha=0,
       reg_lambda=1, scale_pos_weight=1, seed=None, silent=True,
       subsample=1, sum_sample=0.7)

进行模型评测

from sklearn.metrics import mean_squared_error

pred=xg_reg.predict(X_test)

rmse=np.sqrt(mean_squared_error(y_test,pred))
rmse

26348.6994479421

logrmse=np.sqrt(mean_squared_error(np.log(y_test),np.log(pred)))
logrmse

0.11996319328290858

调整模型参数

xg_reg2=xgb.XGBRegressor(objective='reg:linear',colsample_bytree=0.6,learning_rate=0.01,max_depth=5,alpha=10,n_estimators=3000,subsample=0.7,random_state=123)

xg_reg2.fit(X_train, y_train)

XGBRegressor(alpha=10, base_score=0.5, booster='gbtree', colsample_bylevel=1,
       colsample_bytree=0.6, gamma=0, importance_type='gain',
       learning_rate=0.01, max_delta_step=0, max_depth=5,
       min_child_weight=1, missing=None, n_estimators=3000, n_jobs=1,
       nthread=None, objective='reg:linear', random_state=123, reg_alpha=0,
       reg_lambda=1, scale_pos_weight=1, seed=None, silent=True,
       subsample=0.7)

pred2=xg_reg2.predict(X_test)
logrmse=np.sqrt(mean_squared_error(np.log(y_test),np.log(pred2)))
logrmse

0.10447489682428332

pd.get_dummies(train1).shape

(1460, 303)

GridSearch

# gs=GridSearch(xg_reg,{
#     "n_estimators":[100,500,1000,3000],
#     "alpha":[0.01,0.1,1.0,10]
#     "lambda":[...]
# })

Randomized Search

m*n 不重要的点过滤掉

# rs=RandomizedSearch(xg_reg,{
#     "n_estimators":[100,500,1000,3000],
#     "alpha":np.norm(1.0,0.7),
#     "lambda":[...]
# })

交叉验证

params = {"objective":"reg:linear",'colsample_bytree': 0.7,'learning_rate': 0.1,
          'max_depth': 5, 'alpha': 10}

#xgboost可以接受的Data的一种压缩后的数据结构,.lmdb/.h5,基于此可以对训练集的读写进行优化提升训练速度
matrix=xgb.DMatrix(data=X,label=y)

# 本身数据不够用，用10折交叉
# 没必要配测试集了，用量数据进行训练和评测，再平均
X.shape

(1460, 303)

cv_results=xgb.cv(dtrain=matrix,params=params,nfold=10,num_boost_round=500,
                  metrics='rmse',as_pandas=True, verbose_eval=False)
#决策树的学习过程分为2个阶段，分裂和剪枝(前剪枝和后减枝) 都是为了防止过拟合
# verbose_eval把日志关了

type(cv_results)

pandas.core.frame.DataFrame

X=pd.get_dummies(train1) # Convert categorical variable into dummy/indicator variables
# 进行

cv_results.tail()

	test-rmse-mean	test-rmse-std	train-rmse-mean	train-rmse-std
495	26897.374219	6992.056793	1675.321265	103.741632
496	26897.469727	6991.416821	1667.588147	104.618191
497	26897.067188	6991.073168	1661.767676	105.295643
498	26897.811133	6991.131930	1656.616650	105.007822
499	26897.296680	6991.362384	1650.411279	106.048964