python房价数据挖掘_Python数据分析及可视化实例之帝都房价预测

数据集下载易一网络科技 - 付费文章​www.intumu.com

加载数据

import pandas as pd

df=pd.read_csv("MYUNOI.csv") # "石景山","丰台","朝阳","海淀","东城","西城"

df.head()

dist roomnum halls AREA floor subway school price 0 chaoyang 1 0 46.06 middle 1 0 48850 1 chaoyang 1 1 59.09 middle 1 0 46540 2 haidian 5 2 278.95 high 1 1 71662 3 haidian 3 2 207.00 high 1 1 57972 4 fengtai 2 1 53.32 low 1 1 71268

import warnings

warnings.filterwarnings("ignore")

是否有空值

df.isnull().any()

dist False

roomnum False

halls False

AREA False

floor False

subway False

school False

price False

dtype: bool

数据分割

X0, y = df.iloc[:,1:-1],df.iloc[:,-1]

from sklearn.preprocessing import OrdinalEncoder, LabelEncoder,OneHotEncoder

# 对非数字列进行编码

X = OrdinalEncoder().fit_transform(X0)

# X = OneHotEncoder().fit_transform(X) #http://localhost:8888/notebooks/20190819yellowbrick/08-Learning

# y = LabelEncoder().fit_transform(y0)

特征分析

features = X0.columns.tolist()

from yellowbrick.features.importances import FeatureImportances

model = LassoCV(alphas=alphas,cv=5)

viz = FeatureImportances(model, labels=features)

viz.fit(X, y)

viz.poof()

注：学区、地铁，这是必然的！

Alpha参数选取

import numpy as np

from sklearn.linear_model import LassoCV

from yellowbrick.regressor import AlphaSelection

alphas = np.logspace(-10, 1, 400)

# 可视化及验证

model = LassoCV(alphas=alphas,cv=5)

visualizer = AlphaSelection(model)

visualizer.fit(X, y)

visualizer.poof()

使用Alpha参数评估预测结果

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=42)

from sklearn.linear_model import Ridge

from yellowbrick.regressor import PredictionError

visualizer = PredictionError(Ridge(alpha=0.4381))

visualizer.fit(X_train, y_train)

visualizer.score(X_test, y_test)

visualizer.poof()

学习曲线

from sklearn.linear_model import RidgeCV

from yellowbrick.model_selection import LearningCurve

model = RidgeCV()

visualizer = LearningCurve(model, scoring='r2')

visualizer.fit(X, y)

visualizer.poof()

持久化（略）

新手可查阅历史目录：yeayee：Python数据分析及可视化实例目录​zhuanlan.zhihu.com