本文已参与「新人创作礼」活动，一起开启掘金创作之路

import numpy as np
import pandas as pd
from sklearn.linear_model import ElasticNetCV, LassoCV, RidgeCV
from sklearn.ensemble import GradientBoostingRegressor
from sklearn.kernel_ridge import KernelRidge
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import RobustScaler
from sklearn.model_selection import KFold, cross_val_score
from sklearn.metrics import mean_squared_error
from mlxtend.regressor import StackingCVRegressor
import xgboost as xgb
import lightgbm as lgb
import warnings
warnings.filterwarnings('ignore')
import pandas as pd
import numpy as np
import zipfile
from sklearn.model_selection import KFold
import lightgbm as lgb
import seaborn as sns
import matplotlib.pyplot as plt

In [3]:

df=pd.read_excel('去除设定恒定共线性使用标准化.xlsx')

In [4]:

df_train=pd.read_excel('训练验证集.xlsx')

In [5]:

df_test=pd.read_excel('测试集.xlsx')

In [6]:

y_train=df_train.iloc[:,-2:]
x_train=df_train.iloc[:,:-2]
y_test=df_test.iloc[:,-2:]
x_test=df_test.iloc[:,:-2]
y=df.iloc[:,-2:]
x=df.iloc[:,:-2]

In [7]:

# 转换成矩阵
x_train = np.array(x_train)
y_train = np.array(y_train)
x_test = np.array(x_test)
y_test = np.array(y_test)

特征两两相关性分析

In [9]:

plt.figure(figsize=(10,10))
sns.heatmap(df.corr(),annot=True,fmt=".1f",square=True)
plt.show()

plt.figure(figsize=(12,12))
sns.heatmap(df.corr(),vmax=.3,center=0,
            square=True,
           linewidths=.5,
            cbar_kws={"shrink":.5},
            annot=True,
            fmt=".1f")
plt.tight_layout()
plt.show()

sns.pairplot(df)
plt.show()

df

Out[16]:

	挤出机1区（℃）实际	挤出机2区（℃）实际	挤出机3区（℃）实际	挤出机3区（℃）偏差	挤出机4区（℃）实际	挤出机5区（℃）实际	挤出机6区（℃）实际	过渡区1区℃实际	过渡区2区℃实际	过渡区3区℃实际	...	收卷机参数频率实际 (Hz)	收卷机参数电流实际 (A)	热风机参数频率设定 (Hz)	热风机参数频率实际 (Hz)	热风机参数电流实际 (A)	抽吸风机参数频率设定 (Hz)	抽吸风机参数频率实际 (Hz)	抽吸风机参数电流实际 (A)	PFE过滤效率	阻力/pa
0	-0.676828	-0.874057	-0.785047	-0.099374	-0.619479	-0.640760	-0.719819	0.711896	0.355214	-0.387147	...	0.227962	0.225464	0.024059	0.023953	1.212971	0.088767	0.088956	0.983389	95.788889	24.133333
1	0.002809	-0.521390	-0.863245	-0.338083	-0.652002	-0.894566	-0.699741	-1.242728	0.539588	0.819483	...	0.227962	0.225464	0.024059	0.023953	0.924434	0.088767	0.088956	-0.379162	94.700000	23.633333
2	-0.903373	-0.815279	-0.824146	-0.218728	-0.782096	-0.621237	-0.719819	0.467568	-0.382285	-0.588252	...	0.227962	0.225464	0.024059	0.023953	0.635897	0.088767	0.088956	0.302113	95.122222	23.833333
3	0.002809	-0.521390	-0.902343	-0.457438	-0.749573	-0.796949	-0.860365	0.223240	-0.751034	0.216168	...	0.227962	0.225464	0.024059	0.023953	1.212971	0.088767	0.088956	1.664664	95.011111	24.511111
4	-0.450282	-0.815279	-0.902343	-0.457438	-0.814620	-0.835996	-0.639507	-0.998400	0.539588	0.618378	...	0.227962	0.225464	0.024059	0.023953	0.347360	0.088767	0.088956	-1.060438	95.811111	23.633333
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2011	-0.223736	-0.109946	1.150337	-0.159051	0.974172	1.194450	1.227744	-0.021088	-0.566659	0.216168	...	0.227962	0.225464	0.716951	0.717110	-1.095323	0.505909	0.506016	-1.741713	94.541111	22.900000
2012	-0.450282	-0.227502	1.150337	-0.159051	0.974172	1.174926	1.227744	-0.021088	-0.474472	1.020588	...	0.227962	0.225464	0.716951	0.717110	-1.095323	0.505909	0.506016	-1.060438	94.108889	21.300000
2013	-0.450282	-0.227502	1.111238	-0.278406	0.941649	1.194450	1.167510	-0.265416	-0.382285	-0.186042	...	0.227962	0.225464	0.716951	0.717110	-0.806787	0.505909	0.506016	-1.060438	95.563333	22.700000
2014	-0.450282	-0.109946	1.150337	-0.159051	0.974172	1.194450	1.247822	0.223240	-0.474472	0.819483	...	0.227962	0.225464	0.716951	0.717110	-0.806787	0.505909	0.506016	-1.060438	94.924444	22.377778
2015	-0.676828	-0.109946	1.130787	-0.218728	0.974172	1.155403	1.187588	-0.754072	-1.119783	-0.186042	...	0.227962	0.225464	0.716951	0.717110	-0.806787	0.505909	0.506016	-0.379162	94.675556	21.811111

2016 rows × 50 columns

In [17]:

sns.distplot(df["挤出机1区（℃）实际"])
plt.show()

df["挤出机1区（℃）实际"].max()

Out[18]:

4.533722163185907

In [21]:

df["挤出机1区（℃）实际"].unique()

Out[21]:

array([-6.76827577e-01,  2.80934574e-03, -9.03373218e-01, -4.50281936e-01,
       -2.23736295e-01,  2.29354987e-01,  4.55900627e-01,  6.82446268e-01,
        9.08991909e-01, -1.35646450e+00, -1.12991886e+00, -1.58301014e+00,
       -2.26264706e+00, -2.71573834e+00, -2.48919270e+00,  1.13553755e+00,
        1.81517447e+00,  2.04172011e+00,  2.26826575e+00,  2.72135704e+00,
        1.58862883e+00, -1.80955578e+00, -2.94228399e+00, -2.03610142e+00,
        1.36208319e+00,  2.49481140e+00,  2.94790268e+00,  3.17444832e+00,
       -3.39537527e+00, -3.84846655e+00, -4.07501219e+00, -4.52810347e+00,
       -3.62192091e+00, -4.30155783e+00, -4.98119475e+00, -5.43428604e+00,
       -5.66083168e+00, -5.20774039e+00, -4.75464911e+00, -3.16882963e+00,
        3.40099396e+00,  3.62753960e+00,  4.08063088e+00,  4.30717652e+00,
        4.53372216e+00,  3.85408524e+00])

In [23]:

df["挤出机1区（℃）实际"].value_counts()

Out[23]:

 0.002809    340
 0.229355    339
-0.223736    270
 0.455901    220
-0.450282    193
-0.676828    118
 0.682446    107
-0.903373     75
 0.908992     48
-1.129919     28
 1.135538     28
-1.356464     26
 1.362083     25
 1.588629     20
-1.583010     20
 1.815174     17
 2.947903     15
-2.262647     13
 2.268266     12
-2.715738     10
 2.041720      9
-2.489193      8
-2.036101      7
 2.494811      7
-1.809556      6
 3.174448      6
 2.721357      5
-4.075012      5
 4.307177      5
-3.848467      4
 3.627540      4
-2.942284      4
-5.434286      3
-3.395375      3
-5.207740      2
-4.981195      2
-4.301558      2
 3.854085      2
-3.621921      1
-5.660832      1
 4.080631      1
-4.754649      1
-3.168830      1
 4.533722      1
 3.400994      1
-4.528103      1
Name: 挤出机1区（℃）实际, dtype: int64

In [24]:

sns.countplot(x="PFE过滤效率",data=df,palette="bwr")

sns绘图

特征两两相关性分析