一.探索数据
%matplotlib inline
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import seaborn as sns
plt.style.use('fivethirtyeight')
import warnings
warnings.filterwarnings('ignore')
plt.rcParams['font.sans-serif'] = ['Microsoft YaHei']
plt.rcParams['font.serif'] = ['Microsoft YaHei']
from sklearn.datasets.california_housing import fetch_california_housing
housing = fetch_california_housing()
X = housing.data
y = housing.target
X.shape
(20640, 8)
y.shape
(20640,)
housing.feature_names
['MedInc',
'HouseAge',
'AveRooms',
'AveBedrms',
'Population',
'AveOccup',
'Latitude',
'Longitude']
| MedInc | 人均收入 |
|---|---|
| HouseAge | 房龄 |
| AveRooms | 房间数 |
| AveBedrooms | 卧室数 |
| Population | 小区人口数 |
| AveOccup | 房屋居住人数 |
| Longitude | 小区经度 |
| Latitude | 小区纬度 |
housing.feature_names =['人均收入','房龄','房间数', '卧室数', '小区人口数', '房屋居住人数', '小区经度', '小区纬度']
housing.feature_names
['人均收入', '房龄', '房间数', '卧室数', '小区人口数', '房屋居住人数', '小区经度', '小区纬度']
X
array([[ 8.3252 , 41. , 6.98412698, ..., 2.55555556,
37.88 , -122.23 ],
[ 8.3014 , 21. , 6.23813708, ..., 2.10984183,
37.86 , -122.22 ],
[ 7.2574 , 52. , 8.28813559, ..., 2.80225989,
37.85 , -122.24 ],
...,
[ 1.7 , 17. , 5.20554273, ..., 2.3256351 ,
39.43 , -121.22 ],
[ 1.8672 , 18. , 5.32951289, ..., 2.12320917,
39.43 , -121.32 ],
[ 2.3886 , 16. , 5.25471698, ..., 2.61698113,
39.37 , -121.24 ]])
X = pd.DataFrame(X,columns=housing.feature_names)
X.head(10)
| 人均收入 | 房龄 | 房间数 | 卧室数 | 小区人口数 | 房屋居住人数 | 小区经度 | 小区纬度 | |
|---|---|---|---|---|---|---|---|---|
| 0 | 8.3252 | 41.0 | 6.984127 | 1.023810 | 322.0 | 2.555556 | 37.88 | -122.23 |
| 1 | 8.3014 | 21.0 | 6.238137 | 0.971880 | 2401.0 | 2.109842 | 37.86 | -122.22 |
| 2 | 7.2574 | 52.0 | 8.288136 | 1.073446 | 496.0 | 2.802260 | 37.85 | -122.24 |
| 3 | 5.6431 | 52.0 | 5.817352 | 1.073059 | 558.0 | 2.547945 | 37.85 | -122.25 |
| 4 | 3.8462 | 52.0 | 6.281853 | 1.081081 | 565.0 | 2.181467 | 37.85 | -122.25 |
| 5 | 4.0368 | 52.0 | 4.761658 | 1.103627 | 413.0 | 2.139896 | 37.85 | -122.25 |
| 6 | 3.6591 | 52.0 | 4.931907 | 0.951362 | 1094.0 | 2.128405 | 37.84 | -122.25 |
| 7 | 3.1200 | 52.0 | 4.797527 | 1.061824 | 1157.0 | 1.788253 | 37.84 | -122.25 |
| 8 | 2.0804 | 42.0 | 4.294118 | 1.117647 | 1206.0 | 2.026891 | 37.84 | -122.26 |
| 9 | 3.6912 | 52.0 | 4.970588 | 0.990196 | 1551.0 | 2.172269 | 37.84 | -122.25 |
y = pd.DataFrame(y,columns =[ '房价'])
y
| 房价 | |
|---|---|
| 0 | 4.526 |
| 1 | 3.585 |
| 2 | 3.521 |
| 3 | 3.413 |
| 4 | 3.422 |
| ... | ... |
| 20635 | 0.781 |
| 20636 | 0.771 |
| 20637 | 0.923 |
| 20638 | 0.847 |
| 20639 | 0.894 |
20640 rows × 1 columns
data = pd.concat([X,y],axis = 1)
data.head()
| 人均收入 | 房龄 | 房间数 | 卧室数 | 小区人口数 | 房屋居住人数 | 小区经度 | 小区纬度 | 房价 | |
|---|---|---|---|---|---|---|---|---|---|
| 0 | 8.3252 | 41.0 | 6.984127 | 1.023810 | 322.0 | 2.555556 | 37.88 | -122.23 | 4.526 |
| 1 | 8.3014 | 21.0 | 6.238137 | 0.971880 | 2401.0 | 2.109842 | 37.86 | -122.22 | 3.585 |
| 2 | 7.2574 | 52.0 | 8.288136 | 1.073446 | 496.0 | 2.802260 | 37.85 | -122.24 | 3.521 |
| 3 | 5.6431 | 52.0 | 5.817352 | 1.073059 | 558.0 | 2.547945 | 37.85 | -122.25 | 3.413 |
| 4 | 3.8462 | 52.0 | 6.281853 | 1.081081 | 565.0 | 2.181467 | 37.85 | -122.25 | 3.422 |
data.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 20640 entries, 0 to 20639
Data columns (total 9 columns):
人均收入 20640 non-null float64
房龄 20640 non-null float64
房间数 20640 non-null float64
卧室数 20640 non-null float64
小区人口数 20640 non-null float64
房屋居住人数 20640 non-null float64
小区经度 20640 non-null float64
小区纬度 20640 non-null float64
房价 20640 non-null float64
dtypes: float64(9)
memory usage: 1.4 MB
data.describe()
data.hist(bins=40,figsize=(12,12))
plt.show()
ax = sns.heatmap(X.corr(),annot=True,linewidths=0.2,annot_kws={'size':8})
fig=plt.gcf()
fig.set_size_inches(8,8)
plt.xticks(fontsize=8)
plt.yticks(fontsize=8)
bottom, top = ax.get_ylim()
ax.set_ylim(bottom + 0.5, top - 0.5)
plt.show()
数据探索在这部分中，我们不会修改任何数据，只是通过各种统计手段帮助我们对这组数据有一个直观的了解，为之后的处理提取信息。
data.isnull().sum()
人均收入 0
房龄 0
房间数 0
卧室数 0
小区人口数 0
房屋居住人数 0
小区经度 0
小区纬度 0
房价 0
dtype: int64
二.数据预处理
- 缺失值
- 异常值
- 重复值
- 数据变换 标准化
- 编码转换
- 主成分分析
三.算法建模
划分训练集 测试集
from sklearn.model_selection import train_test_split
Xtrain,Xtest,Ytrain,Ytest = train_test_split(X,y,test_size=0.2, random_state=40)
1.线性回归
from sklearn.linear_model import LinearRegression
lr = LinearRegression()
lr =lr.fit(Xtrain,Ytrain)
lr.score(Xtest,Ytest)
0.6075794091011189
2.决策树
from sklearn.tree import DecisionTreeRegressor
dtr = DecisionTreeRegressor(random_state=40)
dtr = dtr.fit(Xtrain, Ytrain)
dtr.score(Xtest,Ytest)
0.615018515749356
3.随机森林
from sklearn.ensemble import RandomForestRegressor
rfr = RandomForestRegressor(random_state = 40)
rfr = rfr.fit(Xtrain, Ytrain)
rfr.score(Xtest, Ytest)
0.7748884071228732
4.回归提升树
from sklearn.ensemble import GradientBoostingRegressor
gbr = GradientBoostingRegressor(n_estimators=100,random_state=40)
gbr = gbr.fit(Xtrain, Ytrain)
gbr.score(Xtest, Ytest)
0.7835836333713624
5.AdaBoostRegressor
from sklearn.ensemble import AdaBoostRegressor
abr = AdaBoostRegressor(random_state=40,n_estimators=300,learning_rate=0.1)
abr = abr.fit(Xtrain, Ytrain)
abr.score(Xtest, Ytest)
0.4957214730873699
6.KNN
from sklearn.neighbors import KNeighborsRegressor
knn = KNeighborsRegressor()
knn = knn.fit(Xtrain, Ytrain)
knn.score(Xtest, Ytest)
0.6972888927183265
7.BaggingRegressor
from sklearn.ensemble import BaggingRegressor
bgr =BaggingRegressor()
bgr = bgr.fit(Xtrain, Ytrain)
bgr.score(Xtest, Ytest)
0.7885598734862586
8.Xgboost
import xgboost as xgb
xgb = xgb.XGBRegressor(n_estimators=120,max_depth=7, learning_rate=0.3,objective ='reg:squarederror')
xgb = xgb.fit(Xtrain, Ytrain)
xgb.score(Xtest,Ytest)
0.8305799952865707
def try_different_method(model, method):
model.fit(Xtrain, Ytrain)
score = model.score(Xtest, Ytest)
result = model.predict(Xtest)
plt.figure()
plt.plot(np.arange(len(result)), Ytest, "go-", label="True value")
plt.plot(np.arange(len(result)), result, "ro-", label="Predict value")
plt.title(f"method:{method}---score:{score}")
plt.legend(loc="best")
plt.show()
from sklearn.linear_model import LinearRegression
lr = LinearRegression()
from sklearn.tree import DecisionTreeRegressor
dtr = DecisionTreeRegressor(random_state=40)
from sklearn.ensemble import RandomForestRegressor
rfr = RandomForestRegressor(random_state = 40)
from sklearn.ensemble import GradientBoostingRegressor
gbr = GradientBoostingRegressor(n_estimators=100,random_state=40)
from sklearn.ensemble import AdaBoostRegressor
abr = AdaBoostRegressor(random_state=40,n_estimators=300,learning_rate=0.1)
from sklearn.neighbors import KNeighborsRegressor
knn = KNeighborsRegressor(n_neighbors=6)
from sklearn.ensemble import BaggingRegressor
bgr =BaggingRegressor(random_state=40)
import xgboost as xgb
xgb = xgb.XGBRegressor(n_estimators=120,max_depth=7, learning_rate=0.3,objective ='reg:squarederror',random_state = 40)
regressor = [lr,dtr,rfr,gbr,abr,knn,bgr,xgb]
def different_model(model,name):
model.fit(Xtrain, Ytrain)
score = model.score(Xtest, Ytest)
result = model.predict(Xtest)
print(f'{name}',score)
name = ['lr','dtr','rfr','gbr','abr','knn','bgr','xgb']
for i,j in zip(regressor,name):
different_model(i,j)
lr 0.6075794091011189
dtr 0.615018515749356
rfr 0.7748884071228732
gbr 0.7835836333713624
abr 0.4957214730873699
knn 0.14199477279929695
bgr 0.7735241405394527
xgb 0.8305799952865707
模型调参
from sklearn import tree
dtr = tree.DecisionTreeRegressor(random_state=40)
dtr = dtr.fit(Xtrain, Ytrain)
dtr.score(Xtest,Ytest)
score = []
C = range(1,20)
for i in C:
dtr = tree.DecisionTreeRegressor(max_depth= i #树的最大深度
,min_samples_leaf=13 #最小分支节点
,random_state=40)
dtr = dtr.fit(Xtrain, Ytrain)
score.append(dtr.score(Xtest,Ytest))
plt.plot(C,score)
print('max score:',max(score),'max_depth:',score.index(max(score))+1)
max score: 0.7269488014943908 max_depth: 12
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-IupvFiyM-1592571954638)(output_42_1.png)]
from sklearn.ensemble import RandomForestRegressor
rfr = RandomForestRegressor(random_state = 40)
rfr = rfr.fit(Xtrain, Ytrain)
rfr.score(Xtest, Ytest)
score = []
C = range(140,240,20)
for i in C:
rfr = RandomForestRegressor(n_estimators = i #基评估器
,random_state = 40#随机种子
#,n_jobs = -1 #决定了使用的CPU内核个数，使用更多的内核能使速度增快，而令n_jobs=-1可以调用所有内核
)
rfr = rfr.fit(Xtrain, Ytrain)
score.append(rfr.score(Xtest,Ytest))
plt.plot(C,score)
print('max score:',max(score))
max score: 0.80334868183468
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-7I1Z9r2B-1592571954641)(output_43_1.png)]
rfr = RandomForestRegressor(n_estimators=200,random_state = 40)
rfr = rfr.fit(Xtrain, Ytrain)
rfr.score(Xtest, Ytest)
0.80334868183468
网格搜索
from sklearn.model_selection import GridSearchCV
tree_param_grid = { 'min_samples_split': list((3,6,9)),'n_estimators':list((10,50,100))}
grid = GridSearchCV(RandomForestRegressor(),param_grid=tree_param_grid, cv=5)
grid.fit(Xtrain, Ytrain)
GridSearchCV(cv=5, error_score='raise-deprecating',
estimator=RandomForestRegressor(bootstrap=True, criterion='mse',
max_depth=None,
max_features='auto',
max_leaf_nodes=None,
min_impurity_decrease=0.0,
min_impurity_split=None,
min_samples_leaf=1,
min_samples_split=2,
min_weight_fraction_leaf=0.0,
n_estimators='warn', n_jobs=None,
oob_score=False, random_state=None,
verbose=0, warm_start=False),
iid='warn', n_jobs=None,
param_grid={'min_samples_split': [3, 6, 9],
'n_estimators': [10, 50, 100]},
pre_dispatch='2*n_jobs', refit=True, return_train_score=False,
scoring=None, verbose=0)
grid.best_score_
0.7992116401881113
grid.best_params_
{'min_samples_split': 6, 'n_estimators': 100}
是否一定要标准化
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
X=scaler.fit_transform(X)
X
array([[ 2.34476576, 0.98214266, 0.62855945, ..., -0.04959654,
1.05254828, -1.32783522],
[ 2.33223796, -0.60701891, 0.32704136, ..., -0.09251223,
1.04318455, -1.32284391],
[ 1.7826994 , 1.85618152, 1.15562047, ..., -0.02584253,
1.03850269, -1.33282653],
...,
[-1.14259331, -0.92485123, -0.09031802, ..., -0.0717345 ,
1.77823747, -0.8237132 ],
[-1.05458292, -0.84539315, -0.04021111, ..., -0.09122515,
1.77823747, -0.87362627],
[-0.78012947, -1.00430931, -0.07044252, ..., -0.04368215,
1.75014627, -0.83369581]])
from sklearn.model_selection import train_test_split
Xtrain,Xtest,Ytrain,Ytest = train_test_split(X,y,test_size=0.2, random_state=40)
from sklearn import tree
dtr = tree.DecisionTreeRegressor(random_state=40)
dtr = dtr.fit(Xtrain, Ytrain)
dtr.score(Xtest,Ytest)
0.6145069972322226
from sklearn.preprocessing import MinMaxScaler
X = housing.data
scaler = MinMaxScaler()
X = scaler.fit_transform(X)
from sklearn.model_selection import train_test_split
Xtrain,Xtest,Ytrain,Ytest = train_test_split(X,y,test_size=0.2, random_state=40)
from sklearn import tree
dtr = tree.DecisionTreeRegressor(random_state=40)
dtr = dtr.fit(Xtrain, Ytrain)
dtr.score(Xtest,Ytest)
0.6161125643854056
954
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
