week_15_Random_Forest

---

霉霉数据分析——基于pandas库

---

背景

摘录《Taylor Swift The Whole Story》

目的

我们尝试通过taylor swift的数据集分析人气单曲、具有舞蹈性和活力的单曲、各个特征的相关程度、专辑流行度。

变量描述

popularity：流行度

Danceability: 舞蹈性

Acousticness: 原声，值越高，歌曲的原声性越强对应为不插电

Energy: 歌曲的能量

instrumentalness：乐器表现

liveness：现场感

Loudness: 响度，值越高，歌曲越响亮

Speechiness: 口语化，值越高，歌词越口语化

Valence: 效价（情绪），值越高，这首歌的情绪就越积极正能量

tempo：节奏感

读入数据

import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)

tay = pd.read_csv('spotify_taylorswift.csv')
tay.head()

	Unnamed: 0	name	album	artist	release_date	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
0	0	Tim McGraw	Taylor Swift	Taylor Swift	2006-10-24	232106	49	0.580	0.575	0.491	0.0	0.1210	-6.462	0.0251	0.425	76.009
1	1	Picture To Burn	Taylor Swift	Taylor Swift	2006-10-24	173066	54	0.658	0.173	0.877	0.0	0.0962	-2.098	0.0323	0.821	105.586
2	2	Teardrops On My Guitar - Radio Single Remix	Taylor Swift	Taylor Swift	2006-10-24	203040	59	0.621	0.288	0.417	0.0	0.1190	-6.941	0.0231	0.289	99.953
3	3	A Place in this World	Taylor Swift	Taylor Swift	2006-10-24	199200	49	0.576	0.051	0.777	0.0	0.3200	-2.881	0.0324	0.428	115.028
4	4	Cold As You	Taylor Swift	Taylor Swift	2006-10-24	239013	50	0.418	0.217	0.482	0.0	0.1230	-5.769	0.0266	0.261	175.558

数据预处理

删除第一列索引

del tay[tay.keys()[0]]
tay.head()

	name	album	artist	release_date	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
0	Tim McGraw	Taylor Swift	Taylor Swift	2006-10-24	232106	49	0.580	0.575	0.491	0.0	0.1210	-6.462	0.0251	0.425	76.009
1	Picture To Burn	Taylor Swift	Taylor Swift	2006-10-24	173066	54	0.658	0.173	0.877	0.0	0.0962	-2.098	0.0323	0.821	105.586
2	Teardrops On My Guitar - Radio Single Remix	Taylor Swift	Taylor Swift	2006-10-24	203040	59	0.621	0.288	0.417	0.0	0.1190	-6.941	0.0231	0.289	99.953
3	A Place in this World	Taylor Swift	Taylor Swift	2006-10-24	199200	49	0.576	0.051	0.777	0.0	0.3200	-2.881	0.0324	0.428	115.028
4	Cold As You	Taylor Swift	Taylor Swift	2006-10-24	239013	50	0.418	0.217	0.482	0.0	0.1230	-5.769	0.0266	0.261	175.558

删除空值的样本

tay.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 171 entries, 0 to 170
Data columns (total 15 columns):
 #   Column            Non-Null Count  Dtype  
---  ------            --------------  -----  
 0   name              171 non-null    object 
 1   album             171 non-null    object 
 2   artist            171 non-null    object 
 3   release_date      171 non-null    object 
 4   length            171 non-null    int64  
 5   popularity        171 non-null    int64  
 6   danceability      171 non-null    float64
 7   acousticness      171 non-null    float64
 8   energy            171 non-null    float64
 9   instrumentalness  171 non-null    float64
 10  liveness          171 non-null    float64
 11  loudness          171 non-null    float64
 12  speechiness       171 non-null    float64
 13  valence           171 non-null    float64
 14  tempo             171 non-null    float64
dtypes: float64(9), int64(2), object(4)
memory usage: 20.2+ KB

tay = tay.dropna()
tay.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 171 entries, 0 to 170
Data columns (total 15 columns):
 #   Column            Non-Null Count  Dtype  
---  ------            --------------  -----  
 0   name              171 non-null    object 
 1   album             171 non-null    object 
 2   artist            171 non-null    object 
 3   release_date      171 non-null    object 
 4   length            171 non-null    int64  
 5   popularity        171 non-null    int64  
 6   danceability      171 non-null    float64
 7   acousticness      171 non-null    float64
 8   energy            171 non-null    float64
 9   instrumentalness  171 non-null    float64
 10  liveness          171 non-null    float64
 11  loudness          171 non-null    float64
 12  speechiness       171 non-null    float64
 13  valence           171 non-null    float64
 14  tempo             171 non-null    float64
dtypes: float64(9), int64(2), object(4)
memory usage: 20.2+ KB

人气排序

#用sort_values对人气进行排序，以获得泰勒最受欢迎的歌曲。
populars = tay.sort_values(by='popularity', ascending=False)

populars.head(13)

	name	album	artist	release_date	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
60	Blank Space	1989 (Deluxe)	Taylor Swift	2014-01-01	231826	82	0.760	0.10300	0.703	0.000000	0.0913	-5.412	0.0540	0.5700	95.997
64	Shake It Off	1989 (Deluxe)	Taylor Swift	2014-01-01	219200	80	0.647	0.06470	0.800	0.000000	0.3340	-5.384	0.1650	0.9420	160.078
95	Lover	Lover	Taylor Swift	2019-08-23	221306	80	0.359	0.49200	0.543	0.000016	0.1180	-7.582	0.0919	0.4530	68.534
82	Delicate	reputation	Taylor Swift	2017-11-10	232253	78	0.750	0.21600	0.404	0.000357	0.0911	-10.178	0.0682	0.0499	95.045
106	You Need To Calm Down	Lover	Taylor Swift	2019-08-23	171360	78	0.771	0.00929	0.671	0.000000	0.0637	-5.617	0.0553	0.7140	85.026
94	Cruel Summer	Lover	Taylor Swift	2019-08-23	178426	77	0.552	0.11700	0.702	0.000021	0.1050	-5.707	0.1570	0.5640	169.994
108	ME! (feat. Brendon Urie of Panic! At The Disco)	Lover	Taylor Swift	2019-08-23	193000	77	0.610	0.03300	0.830	0.000000	0.1180	-4.105	0.0571	0.7280	182.162
83	Look What You Made Me Do	reputation	Taylor Swift	2017-11-10	211853	77	0.766	0.20400	0.709	0.000014	0.1260	-6.471	0.1230	0.5060	128.070
86	Getaway Car	reputation	Taylor Swift	2017-11-10	233626	76	0.562	0.00465	0.689	0.000002	0.0888	-6.745	0.1270	0.3510	172.054
150	You Belong With Me (Taylor’s Version)	Fearless (Taylor's Version)	Taylor Swift	2021-04-09	231124	76	0.632	0.06230	0.773	0.000000	0.0885	-4.856	0.0346	0.4740	130.033
100	Paper Rings	Lover	Taylor Swift	2019-08-23	222400	76	0.811	0.01290	0.719	0.000014	0.0742	-6.553	0.0497	0.8650	103.979
81	Don’t Blame Me	reputation	Taylor Swift	2017-11-10	236413	75	0.615	0.10600	0.534	0.000018	0.0607	-6.719	0.0386	0.1930	135.917
61	Style	1989 (Deluxe)	Taylor Swift	2014-01-01	231000	75	0.588	0.00245	0.791	0.002580	0.1180	-5.595	0.0402	0.4870	94.933

提取 TOP 5 单曲

top_5 = populars.head()

#loc是用来访问一组行和列
top_5.loc[:,['name','popularity']]

	name	popularity
60	Blank Space	82
64	Shake It Off	80
95	Lover	80
82	Delicate	78
106	You Need To Calm Down	78

提取舞蹈性较高的单曲

#有一栏danceability - 舞蹈性，我们要过滤掉那些舞蹈性最高的歌曲（即，超过0.75）。
dance_on = tay.loc[(tay.danceability >= 0.75)]
dance_on.loc[:,['name','album','danceability']].head(13)

	name	album	danceability
56	Treacherous - Original Demo Recording	Red (Deluxe Edition)	0.828
59	Welcome To New York	1989 (Deluxe)	0.789
60	Blank Space	1989 (Deluxe)	0.760
68	How You Get The Girl	1989 (Deluxe)	0.765
71	Clean	1989 (Deluxe)	0.815
76	I Wish You Would - Voice Memo	1989 (Deluxe)	0.781
82	Delicate	reputation	0.750
83	Look What You Made Me Do	reputation	0.766
85	Gorgeous	reputation	0.800
96	The Man	Lover	0.777
98	I Think He Knows	Lover	0.897
100	Paper Rings	Lover	0.811
101	Cornelia Street	Lover	0.824

提取活力值较高的单曲

energy = tay.loc[(tay.energy >= 0.5)]

energy = tay.sort_values(by='energy', ascending=False)

energy.loc[:,['name','album','energy']].head(13)

	name	album	energy
26	Haunted	Speak Now (Deluxe Package)	0.944
11	I'm Only Me When I'm With You	Taylor Swift	0.934
24	Better Than Revenge	Speak Now (Deluxe Package)	0.917
152	Tell Me Why (Taylor’s Version)	Fearless (Taylor's Version)	0.909
57	Red - Original Demo Recording	Red (Deluxe Edition)	0.902
38	Red	Red (Deluxe Edition)	0.896
65	I Wish You Would	1989 (Deluxe)	0.893
74	New Romantics	1989 (Deluxe)	0.889
1	Picture To Burn	Taylor Swift	0.877
163	The Other Side Of The Door (Taylor’s Version)	Fearless (Taylor's Version)	0.873
21	The Story Of Us	Speak Now (Deluxe Package)	0.855
62	Out Of The Woods	1989 (Deluxe)	0.841
108	ME! (feat. Brendon Urie of Panic! At The Disco)	Lover	0.830

查询你最喜欢的专辑

比如，我喜欢《Speak Now》 😃

tay.loc[tay.album == 'Speak Now (Deluxe Package)']

	name	album	artist	release_date	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
15	Mine - POP Mix	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	230546	45	0.696	0.004610	0.768	0.000001	0.1010	-3.863	0.0308	0.692	121.050
16	Sparks Fly	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	260946	50	0.608	0.038700	0.785	0.000000	0.1580	-2.976	0.0311	0.376	114.985
17	Back To December	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	293040	50	0.517	0.020200	0.606	0.000000	0.3240	-5.797	0.0289	0.296	141.929
18	Speak Now	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	240773	49	0.708	0.101000	0.601	0.000000	0.0979	-3.750	0.0306	0.742	118.962
19	Dear John	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	403887	48	0.583	0.183000	0.468	0.000002	0.1110	-5.378	0.0278	0.126	119.375
20	Mean	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	237746	48	0.570	0.445000	0.747	0.000000	0.2190	-3.978	0.0426	0.808	164.004
21	The Story Of Us	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	265636	48	0.575	0.000315	0.855	0.001610	0.0419	-4.827	0.0467	0.840	139.920
22	Never Grow Up	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	290480	44	0.715	0.829000	0.308	0.000000	0.1600	-8.829	0.0305	0.547	124.899
23	Enchanted	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	352200	51	0.535	0.071600	0.618	0.000388	0.1690	-3.913	0.0273	0.228	81.975
24	Better Than Revenge	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	217173	49	0.519	0.016700	0.917	0.000021	0.3590	-3.185	0.0887	0.652	145.882
25	Innocent	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	302266	44	0.553	0.202000	0.604	0.000000	0.1250	-5.295	0.0258	0.186	133.989
26	Haunted	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	242093	47	0.434	0.076900	0.944	0.000000	0.1510	-2.641	0.0581	0.361	162.020
27	Last Kiss	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	367146	47	0.359	0.581000	0.329	0.000037	0.0979	-9.531	0.0293	0.208	84.358
28	Long Live	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	317960	47	0.412	0.042600	0.682	0.000075	0.1060	-4.319	0.0339	0.146	203.959
29	Ours	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	238173	55	0.610	0.505000	0.556	0.000000	0.0851	-7.369	0.0285	0.192	159.838
30	If This Was A Movie	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	234546	52	0.515	0.154000	0.724	0.000004	0.2690	-3.498	0.0267	0.257	147.788
31	Superman	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	276266	47	0.582	0.018700	0.817	0.000002	0.1010	-3.718	0.0337	0.547	131.983
32	Back To December - Acoustic	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	292533	59	0.541	0.731000	0.451	0.000000	0.1970	-6.522	0.0270	0.333	141.713
33	Haunted - Acoustic Version	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	217626	48	0.574	0.841000	0.462	0.000000	0.2800	-5.124	0.0252	0.314	80.858
34	Mine	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	230773	64	0.621	0.003270	0.780	0.000005	0.1840	-2.934	0.0297	0.672	121.038
35	Back To December	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	293040	43	0.525	0.113000	0.676	0.000000	0.2940	-4.684	0.0294	0.281	141.950
36	The Story Of Us	Speak Now (Deluxe Package)	Taylor Swift	2010-01-01	266480	59	0.546	0.004870	0.809	0.000372	0.0437	-3.621	0.0410	0.649	139.910

分析

查看专辑流行度

#"Taylor Swift","Fearless","Speak Now","Red","1989","Reputation","Lover","Folklore","Evermore"
#59,76,64,72,82,78,80,65,72
#'lightblue','gold','purple','red','tan','black','pink','grey','brown'

album = ["Taylor Swift","Fearless","Speak Now","Red","1989","Reputation","Lover","Folklore","Evermore"]
maxpopularity = [59,76,64,72,82,78,80,65,72]
newcolors = ['lightblue','gold','purple','red','tan','black','pink','grey','brown']

plt.figure( figsize = (8,6))
plt.bar(album, maxpopularity,color = newcolors)
plt.title(' Popularity of albums')
plt.xlabel('ALBUM')
plt.ylabel('POPULARITY')
plt.xticks(fontsize=6)
plt.show()

流行度直方图

sns.displot(x='popularity', data=tay, kde=True, color='#a70ad5')
plt.title('Popularity Distribution')

Text(0.5, 1.0, 'Popularity Distribution')

以活力值和舞蹈性为例

import matplotlib.pyplot as plt
import numpy as np

energyy = tay.energy
dancee = tay.danceability

correlation_length = energyy.corr(dancee)
print(correlation_length)
ax1 = tay.plot.scatter(x = 'energy',y = 'danceability',c = 'red')

0.06266927386034048

Energy 和 Acousticness的相关性

sns.set_theme(color_codes=True)
plt.figure(figsize=(18, 6))
sns.regplot(x='acousticness', y='energy', data=tay)
plt.title('Acousticness vs Energy')

Text(0.5, 1.0, 'Acousticness vs Energy')

Energy 和 Loudness相关性

sns.set_theme(color_codes=True)
plt.figure(figsize=(18, 6))
sns.regplot(x='energy', y='loudness', data=tay)
plt.title('Energy vs Loudness')

Text(0.5, 1.0, 'Energy vs Loudness')

相关系数矩阵

tay.corr()

C:\Users\dogfa\AppData\Local\Temp\ipykernel_10964\3121711199.py:1: FutureWarning: The default value of numeric_only in DataFrame.corr is deprecated. In a future version, it will default to False. Select only valid columns or specify the value of numeric_only to silence this warning.
  tay.corr()

	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
length	1.000000	0.011772	-0.301561	0.038749	-0.114792	-0.081288	-0.148412	0.044126	-0.414447	-0.420405	0.010425
popularity	0.011772	1.000000	0.072622	-0.117842	0.127495	0.035638	-0.406730	0.122576	-0.478262	0.034154	-0.015669
danceability	-0.301561	0.072622	1.000000	-0.143085	0.062669	-0.051837	-0.015766	0.002587	0.183860	0.379786	-0.235370
acousticness	0.038749	-0.117842	-0.143085	1.000000	-0.710055	0.140655	-0.065387	-0.736624	0.143127	-0.231232	-0.134467
energy	-0.114792	0.127495	0.062669	-0.710055	1.000000	0.000281	0.046364	0.784973	-0.179336	0.490371	0.209914
instrumentalness	-0.081288	0.035638	-0.051837	0.140655	0.000281	1.000000	-0.059132	-0.084224	-0.029729	0.020076	0.043274
liveness	-0.148412	-0.406730	-0.015766	-0.065387	0.046364	-0.059132	1.000000	0.016324	0.357924	-0.017264	0.034934
loudness	0.044126	0.122576	0.002587	-0.736624	0.784973	-0.084224	0.016324	1.000000	-0.409577	0.299926	0.171503
speechiness	-0.414447	-0.478262	0.183860	0.143127	-0.179336	-0.029729	0.357924	-0.409577	1.000000	0.120352	-0.027812
valence	-0.420405	0.034154	0.379786	-0.231232	0.490371	0.020076	-0.017264	0.299926	0.120352	1.000000	-0.006056
tempo	0.010425	-0.015669	-0.235370	-0.134467	0.209914	0.043274	0.034934	0.171503	-0.027812	-0.006056	1.000000

相关系数矩阵图

import seaborn as sns
import matplotlib.pyplot as plt

sns.heatmap(tay.corr())

C:\Users\dogfa\AppData\Local\Temp\ipykernel_10964\54934055.py:1: FutureWarning: The default value of numeric_only in DataFrame.corr is deprecated. In a future version, it will default to False. Select only valid columns or specify the value of numeric_only to silence this warning.
  sns.heatmap(tay.corr())





<Axes: >

相关性矩阵图

sns.pairplot(tay[2:])

<seaborn.axisgrid.PairGrid at 0x29d8b437b50>

Random Forest Regression

算法原理

随机森林中的每棵树在称为自助聚集 (bagging) 的过程中随机对训练数据子集进行抽样。该模型适合这些较小的数据集，并汇总预测结果。通过有放回抽样，可以重复使用同一数据的几个实例，结果就是，这些树不仅基于不同的数据集进行训练，而且还使用不同的特性做出决策。

整理数据

# 删除album列
dataset = tay.drop(['album','artist','release_date','name'], axis=1)

dataset

	length	popularity	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
0	232106	49	0.580	0.575	0.491	0.000000	0.1210	-6.462	0.0251	0.425	76.009
1	173066	54	0.658	0.173	0.877	0.000000	0.0962	-2.098	0.0323	0.821	105.586
2	203040	59	0.621	0.288	0.417	0.000000	0.1190	-6.941	0.0231	0.289	99.953
3	199200	49	0.576	0.051	0.777	0.000000	0.3200	-2.881	0.0324	0.428	115.028
4	239013	50	0.418	0.217	0.482	0.000000	0.1230	-5.769	0.0266	0.261	175.558
...	...	...	...	...	...	...	...	...	...	...	...
166	277591	74	0.660	0.162	0.817	0.000000	0.0667	-6.269	0.0521	0.714	135.942
167	244236	65	0.609	0.849	0.373	0.000000	0.0779	-8.819	0.0263	0.130	106.007
168	189495	67	0.588	0.225	0.608	0.000000	0.0920	-7.062	0.0365	0.508	90.201
169	208608	66	0.563	0.514	0.473	0.000012	0.1090	-11.548	0.0503	0.405	101.934
170	242157	64	0.624	0.334	0.624	0.000000	0.0995	-7.860	0.0539	0.527	80.132

171 rows × 11 columns

x = dataset.drop(['popularity'],axis=1)
x

	length	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
0	232106	0.580	0.575	0.491	0.000000	0.1210	-6.462	0.0251	0.425	76.009
1	173066	0.658	0.173	0.877	0.000000	0.0962	-2.098	0.0323	0.821	105.586
2	203040	0.621	0.288	0.417	0.000000	0.1190	-6.941	0.0231	0.289	99.953
3	199200	0.576	0.051	0.777	0.000000	0.3200	-2.881	0.0324	0.428	115.028
4	239013	0.418	0.217	0.482	0.000000	0.1230	-5.769	0.0266	0.261	175.558
...	...	...	...	...	...	...	...	...	...	...
166	277591	0.660	0.162	0.817	0.000000	0.0667	-6.269	0.0521	0.714	135.942
167	244236	0.609	0.849	0.373	0.000000	0.0779	-8.819	0.0263	0.130	106.007
168	189495	0.588	0.225	0.608	0.000000	0.0920	-7.062	0.0365	0.508	90.201
169	208608	0.563	0.514	0.473	0.000012	0.1090	-11.548	0.0503	0.405	101.934
170	242157	0.624	0.334	0.624	0.000000	0.0995	-7.860	0.0539	0.527	80.132

171 rows × 10 columns

y = dataset['popularity']
y

0      49
1      54
2      59
3      49
4      50
       ..
166    74
167    65
168    67
169    66
170    64
Name: popularity, Length: 171, dtype: int64

划分train test

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=1)

x_train

	length	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
89	230373	0.719	0.032900	0.469	0.000000	0.1690	-8.792	0.0533	0.0851	120.085
88	211506	0.624	0.060400	0.691	0.000011	0.1380	-6.686	0.1960	0.2840	160.024
165	220839	0.599	0.816000	0.494	0.000000	0.1010	-7.610	0.0372	0.4400	142.893
110	293453	0.557	0.808000	0.496	0.000173	0.0772	-9.602	0.0563	0.2650	149.983
48	284866	0.624	0.632000	0.340	0.033700	0.0805	-12.411	0.0290	0.2610	129.987
...	...	...	...	...	...	...	...	...	...	...
133	215626	0.546	0.418000	0.613	0.000000	0.1030	-7.589	0.0264	0.5350	79.015
137	260440	0.515	0.855000	0.545	0.000020	0.0921	-9.277	0.0353	0.5350	88.856
72	245560	0.422	0.049300	0.692	0.000026	0.1770	-5.447	0.0549	0.1970	184.014
140	257773	0.535	0.876000	0.561	0.000136	0.1150	-11.609	0.0484	0.2870	96.103
37	295720	0.588	0.000197	0.825	0.001380	0.0884	-5.882	0.0328	0.3970	129.968

136 rows × 10 columns

x_test

	length	danceability	acousticness	energy	instrumentalness	liveness	loudness	speechiness	valence	tempo
92	235466	0.661	0.921000	0.151	0.000000	0.1300	-12.864	0.0354	0.230	94.922
113	231000	0.688	0.481000	0.653	0.004140	0.1060	-8.558	0.0403	0.701	147.991
19	403887	0.583	0.183000	0.468	0.000002	0.1110	-5.378	0.0278	0.126	119.375
69	250093	0.481	0.678000	0.435	0.000000	0.0928	-8.795	0.0321	0.107	143.950
53	286613	0.619	0.187000	0.506	0.000015	0.1010	-7.327	0.0315	0.274	126.030
161	237338	0.476	0.040600	0.564	0.000000	0.1020	-5.677	0.0269	0.167	143.929
108	193000	0.610	0.033000	0.830	0.000000	0.1180	-4.105	0.0571	0.728	182.162
14	179066	0.459	0.040200	0.753	0.000000	0.0863	-3.827	0.0537	0.483	199.997
99	234146	0.662	0.028000	0.747	0.006150	0.1380	-6.926	0.0736	0.487	150.088
107	223293	0.756	0.130000	0.449	0.000000	0.1140	-8.746	0.0344	0.399	111.011
11	213053	0.563	0.004520	0.934	0.000807	0.1030	-3.629	0.0646	0.518	143.964
4	239013	0.418	0.217000	0.482	0.000000	0.1230	-5.769	0.0266	0.261	175.558
117	208906	0.602	0.888000	0.494	0.000026	0.0902	-10.813	0.0277	0.374	94.955
42	232120	0.661	0.002150	0.729	0.001300	0.0477	-6.561	0.0376	0.668	103.987
122	237266	0.593	0.670000	0.700	0.000007	0.1160	-9.016	0.0492	0.451	141.898
125	234000	0.644	0.916000	0.284	0.000015	0.0909	-12.879	0.0821	0.328	150.072
147	235766	0.627	0.130000	0.792	0.000004	0.0845	-4.311	0.0310	0.415	119.054
35	293040	0.525	0.113000	0.676	0.000000	0.2940	-4.684	0.0294	0.281	141.950
81	236413	0.615	0.106000	0.534	0.000018	0.0607	-6.719	0.0386	0.193	135.917
31	276266	0.582	0.018700	0.817	0.000002	0.1010	-3.718	0.0337	0.547	131.983
51	220600	0.649	0.021300	0.777	0.000335	0.2100	-5.804	0.0406	0.587	126.018
75	216333	0.592	0.829000	0.128	0.000000	0.5270	-17.932	0.5890	0.150	78.828
78	208186	0.613	0.052700	0.764	0.000000	0.1970	-6.509	0.1360	0.417	160.015
73	267106	0.474	0.707000	0.480	0.000108	0.0903	-8.894	0.0622	0.319	170.109
40	219720	0.622	0.004540	0.469	0.000002	0.0335	-6.798	0.0363	0.679	77.019
84	227906	0.574	0.122000	0.610	0.000001	0.1300	-7.283	0.0732	0.374	74.957
47	202960	0.627	0.016200	0.816	0.002080	0.0965	-6.698	0.0774	0.648	157.043
29	238173	0.610	0.505000	0.556	0.000000	0.0851	-7.369	0.0285	0.192	159.838
16	260946	0.608	0.038700	0.785	0.000000	0.1580	-2.976	0.0311	0.376	114.985
105	200306	0.739	0.736000	0.320	0.000147	0.1110	-10.862	0.2390	0.351	79.970
85	209680	0.800	0.071300	0.535	0.000009	0.2130	-6.684	0.1350	0.451	92.027
154	243136	0.402	0.003300	0.732	0.000000	0.1080	-4.665	0.0484	0.472	161.032
157	279359	0.499	0.000191	0.815	0.000000	0.1810	-4.063	0.0341	0.344	95.999
5	207106	0.589	0.004910	0.805	0.000000	0.2400	-4.055	0.0293	0.591	112.982
94	178426	0.552	0.117000	0.702	0.000021	0.1050	-5.707	0.1570	0.564	169.994

y_train

89     71
88     68
165    63
110    68
48     59
       ..
133    65
137    65
72     65
140    63
37     62
Name: popularity, Length: 136, dtype: int64

y_test

92     67
113    63
19     48
69     60
53     58
161    61
108    77
14     48
99     70
107    74
11     50
4      50
117    62
42     64
122    61
125    60
147    74
35     43
81     75
31     47
51     58
75      0
78     74
73     62
40     65
84     66
47     61
29     55
16     50
105    68
85     72
154    74
157    61
5      47
94     77
Name: popularity, dtype: int64

训练模型

from sklearn.ensemble import RandomForestRegressor
regressor = RandomForestRegressor(n_estimators=100, random_state=27)
regressor.fit(x_train, y_train)

RandomForestRegressor(random_state=27)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

计算准确率

regressor.score(x_test,y_test)

0.41401740646745133

查看预测值和实际值

compare = pd.DataFrame(y_test)
compare.columns = ['actual']
compare

	actual
92	67
113	63
19	48
69	60
53	58
161	61
108	77
14	48
99	70
107	74
11	50
4	50
117	62
42	64
122	61
125	60
147	74
35	43
81	75
31	47
51	58
75	0
78	74
73	62
40	65
84	66
47	61
29	55
16	50
105	68
85	72
154	74
157	61
5	47
94	77

y_pred = regressor.predict(x_test)

compare['predict'] = y_pred.round(0)
compare

	actual	predict
92	67	60.0
113	63	66.0
19	48	55.0
69	60	63.0
53	58	61.0
161	61	53.0
108	77	59.0
14	48	60.0
99	70	66.0
107	74	61.0
11	50	55.0
4	50	54.0
117	62	63.0
42	64	68.0
122	61	65.0
125	60	68.0
147	74	57.0
35	43	55.0
81	75	68.0
31	47	56.0
51	58	64.0
75	0	43.0
78	74	69.0
73	62	67.0
40	65	65.0
84	66	68.0
47	61	67.0
29	55	62.0
16	50	57.0
105	68	68.0
85	72	72.0
154	74	60.0
157	61	58.0
5	47	56.0
94	77	67.0