Titanic Missing Value Handling & Survival Prediction

1. Kaggle Titanic data titanic complete download, shared by the original author's conscience

https://download.csdn.net/download/lansui7312/9936840

 

2. Missing value handling

# -*- coding:utf -
import pandas as pd
import numpy as np
from sklearn.ensemble import RandomForestRegressor

df = pd.read_csv('train.csv',header=0)
# SibSp number of cousins/sisters
# Parch number of parents and children
# Cabin
#Embarked Port of Embarkation

# df.head(20)
# df.info()
# print(df.describe())
# print(df.columns[1])


# Male and Female Survival Rate View
x=[df[(df.Sex=='male')]['Sex'].size,df[(df.Sex=='female')]['Sex'].size]
# print(x)
y=[df[(df.Sex=='male') & (df.Survived==1)]['Sex'].size,
   df[(df.Sex=='female') & (df.Survived==1)]['Sex'].size]

# print(y[1]*100/x[1],y[0]*100/x[0])

# missing value handling
# 1. Delete missing lines
# 2. The boarding location is not important, use the mode to assign the value
df.Embarked[df.Embarked.isnull()] = df.Embarked.dropna().mode().values
# print(df.Embarked)

#3. Nominal attribute, assign missing values, because the missing itself may also be an implicit information. For example, Cabin means no cabin
df.Cabin[df.Cabin.isnull()]='U0'
# print(df.Cabin)

#4. Use regression, random forest, etc. models to predict the value of missing attributes
# Because age is a very important feature, it is necessary to ensure a certain accuracy of missing value filling
age_df=df[['Age','Survived','Fare','Parch','SibSp','Pclass']]
age_df_notnull = age_df.loc[(df.Age.notnull())]
age_df_isnull = age_df.loc[(df.Age.isnull())]
X = age_df_notnull.values[:,1:]
Y = age_df_notnull.values[:,0]
# y1 = age_df_notnull.values[:,2:]
# X[:,0] is a way of writing arrays in numpy. For a two-dimensional array, take all the data in the first dimension of the two-dimensional array, and take the 0th data in the second dimension. Intuitively, X [:,0] is to take the 0th data of all rows, X[:,1] is to take the first data of all rows
# print(X)
rfr = RandomForestRegressor(n_estimators=1000,n_jobs=-1)
rfr.fit(X,Y)
predictAges = rfr.predict(age_df_isnull.values[:,1:])
df.loc[(df.Age.isnull()),'Age'] = predictAges

print(df.describe())

 

3.