对泰坦尼克号的数据 进行特征分析、数据清理、数据填充、处理分类特征、将连续特征转化为离散特征、合并特征、制作模型、模型预测

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

import warnings
warnings.filterwarnings('ignore')

train_df=pd.read_csv('train.csv')

test_df=pd.read_csv('test.csv')
combine=[train_df,test_df]

PassengerId=test_df['PassengerId']

特征分析
分析存活率与几个因素之间的关系

train_df[['Pclass','Survived']].groupby(['Pclass'],as_index=False).mean().sort_values(by='Survived',ascending=False)

train_df[['Sex','Survived']].groupby(['Sex'],as_index=False).mean().sort_values(by='Survived',ascending=False)

train_df[['SibSp','Survived']].groupby(['SibSp'],as_index=False).mean().sort_values(by='Survived',ascending=False)

train_df[['Parch','Survived']].groupby(['Parch'],as_index=False).mean().sort_values(by='Survived',ascending=False)

train_df[['Embarked','Survived']].groupby(['Embarked'],as_index=False).mean().sort_values(by='Survived',ascending=False)

fig=plt.figure(figsize=(12,8))

ax1=fig.add_subplot(1,2,1)
ax2=fig.add_subplot(1,2,2)

age1=train_df.Age[train_df['Survived']==0]
age2=train_df.Age[train_df['Survived']==1]

ax1.set_title('Survived=0')
ax1.hist(age1,bins=20)
ax1.set_xlabel('age')
ax1.set_yticks([10,20,30,40,50,60])

ax2.set_title('Survived=1')
ax2.hist(age2,bins=20)
ax2.set_xlabel('age')
ax2.set_yticks([10,20,30,40,50,60])

plt.show()

fare=train_df.Fare[train_df['Survived']==1]
plt.hist(fare)
plt.show()

数据清理

train_df=train_df.drop(['Ticket','Cabin'],axis=1)
test_df=test_df.drop(['Ticket','Cabin'],axis=1)
combine=[train_df,test_df]
train_df.head()

train_df.Name.str.extract(' ([a-zA-Z]+)\.',expand=False)

for database in combine:
    database['Title']=database.Name.str.extract(' ([a-zA-Z]+)\.',expand=False)

pd.crosstab(train_df['Title'],train_df['Sex'])

for database in combine:
    database['Title']=database['Title'].replace(['Lady','Countess','Capt','Col','Don','Dr','Major','Rev','Sir','Jonkheer','Dona'],'Rare')

    database['Title']=database['Title'].replace('Mlle','Miss')

    database['Title']=database['Title'].replace('Ms','Miss')
    database['Title']=database['Title'].replace('Mme','Mrs')

train_df[['Title','Survived']].groupby(['Title'],as_index=False).mean().sort_values(by='Survived',ascending=False)

train_df.drop(['PassengerId','Name'],axis=1,inplace=True)
test_df.drop(['PassengerId','Name'],axis=1,inplace=True)
combine=[train_df,test_df]
train_df

数据填充

train_df.info()

从图中会发现 Age 和 Embarked 都有缺失值

freq=train_df.Embarked.dropna().mode()[0]

for database in combine:
    database.Embarked=database.Embarked.fillna(freq)

grp = train_df.groupby(['Pclass','Sex','Title'])['Age'].mean().reset_index()

def fill_age(x):
    return grp[(grp['Pclass']==x['Pclass']) & (grp['Sex']==x['Sex']) & (grp['Title']==x['Title'])].Age.values[0]

train_df['Age'] = train_df.apply(lambda x: fill_age(x) if np.isnan(x['Age']) else x['Age'] ,axis=1)
test_df['Age'] = test_df.apply(lambda x: fill_age(x) if np.isnan(x['Age']) else x['Age'] ,axis=1)
combine = [train_df,test_df]

处理分类特征
将数据根据分类转化为数字 0.1.2.3...

for database in combine:
    database['Sex'] = database['Sex'].map({'female':1,'male':0}).astype(int)

title_mapping = {'Mr':1 ,'Miss':2 ,'Mrs':3 ,'Master':4 ,'Rare':5}
for database in combine:
    database['Title'] = database['Title'].map(title_mapping).astype(int)

将连续特征转化为离散特征
将连续的数据 分为几组\几类

train_df['AgeBand'] =pd.cut(train_df['Age'],5)
train_df['AgeBand']

通过切割数据 可以将Age准确的分为五个组 并将不同组用1.2.3.4.5代替

for database in combine:
    database.loc[ database['Age'] <16,'Age']=0
    database.loc[(database['Age'] >=16)&(database['Age'] <32),'Age']=1
    database.loc[(database['Age'] >=32)&(database['Age'] <48),'Age']=2
    database.loc[(database['Age'] >=48)&(database['Age'] <64),'Age']=3
    database.loc[ database['Age'] >=64,'Age']=4

train_df.drop(['AgeBand'],axis=1,inplace=True)
combine=[train_df,test_df]
train_df

利用qcut处理Fare的数据

train_df['FareBand']=pd.qcut(train_df['Fare'],4,duplicates='drop')

test_df['Fare'] = test_df['Fare'].fillna(test_df['Fare'].mean())

for dataset in combine:
    dataset.loc[ dataset['Fare'] 7.91,'Fare']=0
    dataset.loc[(dataset['Fare'] >7.91)&(dataset['Fare'] 14.454),'Fare']=1
    dataset.loc[(dataset['Fare'] >14.454)&(dataset['Fare'] 31),'Fare']=2
    dataset.loc[dataset['Fare'] >31,'Fare']=3
    dataset['Fare'] = dataset['Fare'].astype(int)

train_df.drop(['FareBand'],axis=1,inplace=True)
combine=[train_df,test_df]
train_df

合并特征

将SibSp(兄弟姐妹配偶)和Parch(父母)两组数据合并为一组Family
后借助family判断是否为一个人出行IsAlong

for database in combine:
    database['Family']=database['SibSp'] + database['Parch'] + 1

for database in combine:
    database['IsAlong']=1
    database.loc[database['Family']>1,'IsAlong']=0

train_df.drop(['SibSp','Parch','Family'],inplace=True,axis=1)
test_df.drop(['SibSp','Parch','Family'],inplace=True,axis=1)

train_df

制作模型

X_train = train_df.drop('Survived',axis=1)
Y_train = train_df['Survived']

from sklearn.linear_model import LogisticRegression

logreg = LogisticRegression()
logreg.fit(X_train,Y_train)
Y_pred = logreg.predict(test_df)
acc_log = round(logreg.score(X_train,Y_train)*100)

from sklearn.svm import SVC

svc = SVC()
svc.fit(X_train,Y_train)
Y_prd = svc.predict(test_df)
acc_svc = round(svc.score(X_train,Y_train)*100,2)

from sklearn.tree import DecisionTreeClassifier

decision_tree = DecisionTreeClassifier()
decision_tree.fit(X_train,Y_train)
Y_pred = decision_tree.predict(test_df)
acc_decision_tree = round(decision_tree.score(X_train,Y_train)*100,2)

from sklearn.ensemble import RandomForestClassifier

random_forest = RandomForestClassifier()
random_forest.fit(X_train,Y_train)
Y_pred = random_forest.predict(test_df)
acc_random_forest = round(random_forest.score(X_train,Y_train)*100,2)

from sklearn.neighbors import KNeighborsClassifier

knn = KNeighborsClassifier()
knn.fit(X_train,Y_train)
Y_pred = knn.predict(test_df)
acc_knn = round(knn.score(X_train,Y_train)*100,2)

from sklearn.naive_bayes import GaussianNB

gaussian = GaussianNB()
gaussian.fit(X_train,Y_train)
Y_pred = gaussian.predict(test_df)
acc_gaussian = round(gaussian.score(X_train,Y_train)*100,2)

将模型根据评分进行排名

models = pd.DataFrame({
    'Modle':['Logistic Regression','Support Vector Machines','KNN','Naive Bayes','Decision Tree','Random Forest'],
    'Score':[acc_log,acc_svc,acc_knn,acc_gaussian,acc_decision_tree,acc_random_forest]
})
models.sort_values(by='Score',ascending=False)

模型预测
有以上模型的排名情况 可选择决策树对数据进行预测

from sklearn.tree import DecisionTreeClassifier

decision_tree = DecisionTreeClassifier()
decision_tree.fit(X_train,Y_train)
Y_pred = decision_tree.predict(test_df)
acc_decision_tree = round(decision_tree.score(X_train,Y_train)*100,2)

submission = pd.DataFrame({
    'PassengerId':PassengerId,
    'Survived':Y_pred
})

submission.to_csv('submission.csv',index=False)
submission

Original: https://blog.csdn.net/An_0330/article/details/121201245
Author: ximu VS code
Title: python 泰坦尼克号存活率分析

相关阅读

Title: pytorch安装详细步骤

文章目录

• （一）win—配置tensorflow-GPU
• （二）安装 pytorch
  *
• 2.1 创建虚拟环境
• 2.2正式安装pytorch
• 2.3 验证是否安装成功
• （三）本文参考链接如下：（感谢各位大佬）

（一）win—配置tensorflow-GPU

直接查看这条链接即可：win-配置tf-GPU
本人用的conda和tensorflow-GPU版本下载：提取码：98ot
环境：win10+anaconda
注：anaconda安装步骤略，以下步骤默认anaconda已安装。

（二）安装 pytorch

2.1 创建虚拟环境

conda create --name pytorch python=3.8.1

注意，这里的 pytorch 是虚拟环境的名称，可随意取。3.8.1是我机器上的python版本，可结合自己安装的python版本灵活变换。

activate pytorch

2.2正式安装pytorch

****：pytorch 官网链接

注意：如果笔记本有独立显卡（NVIDIA）的话，可以如上选择对应的CUDA版本，否则选择CPU。

• 本人采用的方法
  在pytorch虚拟环境下，创建.condarc文件
  
  在虚拟环境中输入如下命令：

conda config --set show_channel_urls yes

• 之后在自己电脑的用户文件下找到一个.condarc文件
  
  用记事本打开这个文件，然后用如下代码代替其中的内容：

channels:
  - defaults
show_channel_urls: true
channel_alias: https://mirrors.tuna.tsinghua.edu.cn/anaconda
default_channels:
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/r
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/pro
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/msys2
custom_channels:
  conda-forge: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  msys2: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  bioconda: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  menpo: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  pytorch: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  simpleitk: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud

记住.condarc配置好后一定要保存。

• 或者使用清华镜像源
  在浏览pytorch安装帮助的相关帖子时有人说清华源停止镜像了，但是现在清华源已经恢复提供镜像了，所以还是可以用的。附上清华源Anaconda镜像使用帮助，大家可以读一下这段帮助（不读也没关系下面给出详细步骤）。

conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch

复制这条命令在 Anaconda Prompt 里输入。

2.3 验证是否安装成功

是否安装成功分两个方面。

• 一个是在prompt里面

（1）在命令行左边为 pytorch 环境中，输入 python

（2）之后，输入 import torch，如果没有报错，意味着 PyTorch 已经顺利安装了。

• 一个是在jupyter notebook里面调用
  首先，在菜单中打开Anaconda Prompt，然后安装插件。
  命令行输入：

conda install nb_conda

然后进入创建的pytorch环境，命令行输入:

conda install ipykernel

安装成功的样子如下：

进入Anaconda中可查看，多了pytorch框架：

（三）本文参考链接如下：（感谢各位大佬）

WIN10下pytorch环境配置（安装了半天的血泪史）
WIn10+Anaconda环境下安装PyTorch(避坑指南)
win10下使用anaconda安装pytorch（清华镜像）
如何让Jupyter Notebook支持pytorch

Original: https://blog.csdn.net/weixin_54546190/article/details/120754242
Author: ☞源仔
Title: pytorch安装详细步骤