泰坦尼克号浅分析

import numpy
import pandas
from matplotlib import pyplot


class Titanic:
    """
	Titanic.csv           根目录的泰坦尼克号的详情文档
    PassengerId       乘客的唯一id
    Survived          1：获救       0：死亡
    Pclass            座舱等级    3最好
    Name              姓名
    Sex               性别
    Age               年龄
    SibSp             船上有没有兄弟姐妹
    Parch             穿上有没有父母等直系亲属
    Ticket            船票
    Fare              花费
    Cabin             座舱号
    Embarked          从那个港口登船

    任务：
    1  统计存活率多少
    2  那个年龄段存活率最高
    3  女性存活率是否高于男性
    4  船上是否存在明显的贫富差距
    5  头等仓的存活率是否高于经济舱
    6  有亲属在船上的乘客比率  有亲属是否影响存活率
    7  从那个港口是否影响存活率
    8  不同年龄段女性的获救率
    """

    def __init__(self):
        self.df_titanic = pandas.read_csv('Titanic.csv')
        self.columns = self.df_titanic.columns
        self.index = self.df_titanic.index
        self.sum_ = self.df_titanic.count()[0]
    def _rate_survival(self,dataFarm):
        cuihuo = dataFarm['Survived'].value_counts()[1]
        zong_shu = dataFarm.count()[0]
        return cuihuo/zong_shu

    # 1  统计存活率多少
    def survival_rate(self):
        return self._rate_survival(self.df_titanic)
        # return survived
        # return self.total

    # 2  那个年龄段存活率最高
    def max_survival_age_bracket(self):
        avg = self.df_titanic['Age'].mean()
        data = self.df_titanic.fillna(avg)
        max_age = data['Age'].max()
        n = int(max_age // 10 + 1)
        # print(data['Age'])
        max_survival_age = 0
        max_rate = 0
        for i in range(n):
            ages = data[(i * 10 < data['Age']) & (data['Age'] <= (i + 1) * 10)]
            sum_ages = ages.count()[0]
            sum_survival = ages['Survived'].value_counts()[1]
            rate = sum_survival / sum_ages
            if rate > max_survival_age:
                max_survival_age = i
                max_rate = rate
        return '{}岁--{}岁的人存活率最大，区间存活率为{}'.format(max_survival_age * 10, (max_survival_age + 1) * 10, max_rate)

    # 3  女性存活率是否高于男性
    def sex_survival(self):
        nv = self.df_titanic[self.df_titanic['Sex'].str.contains('female')]
        nv_survivals_sum = nv['Survived'].value_counts()
        nv_rate = nv_survivals_sum[1] / (nv.count()[0])
        nan = self.df_titanic[self.df_titanic['Sex'].str.contains('male')]
        nan_survivals_sum = nan['Survived'].value_counts()
        nan_rate = nan_survivals_sum[1] / (nan.count()[0])

        if nv_rate > nan_rate:
            return '女士存活率({}) 大于 男士存活率({})'.format(nv_rate, nan_rate)
        return '女士存活率({}) 小于 男士存活率({})'.format(nv_rate, nan_rate)

    # 4  船上是否存在明显的贫富差距
    def rich_gap(self):
        pclass = self.df_titanic['Pclass'].value_counts()
        one_pc_rate = pclass[1] / self.sum_
        two_pc_rate = pclass[2] / self.sum_
        three_pc_rate = pclass[3] / self.sum_
        rate = three_pc_rate / one_pc_rate
        if rate > 1.5:
            return '座舱等级为3的比率{}，座舱等级为2的比率{}，座舱等级为1的比率{}，存在明显的贫富差距'.format(three_pc_rate, two_pc_rate, one_pc_rate)
        else:
            return '座舱等级为3的比率{}，座舱等级为2的比率{}，座舱等级为1的比率{}，不存在明显的贫富差距'.format(three_pc_rate, two_pc_rate, one_pc_rate)

    # 5  头等仓的存活率是否高于经济舱
    def pclass_survivesd(self):
        tou_pclass = self.df_titanic[self.df_titanic['Pclass'] == 3]
        tou_sum = tou_pclass.count()[0]
        tou_rate = tou_pclass['Survived'].value_counts()[1]/tou_sum
        jing_pclass = self.df_titanic[self.df_titanic['Pclass'] == 1]
        jing_sum = jing_pclass.count()[0]
        jing_rate = jing_pclass['Survived'].value_counts()[1]/jing_sum
        if jing_rate > tou_rate:
            return '经济舱{}生存率高于头等舱{}'.format(jing_rate,tou_rate)
        return '头等舱{}生存率高于经济舱{}'.format(tou_rate,jing_rate)
    # 6  有亲属在船上的乘客比率  有亲属是否影响存活率
    def qinshu_survived_ziji(self):
        wu_peoples = self.df_titanic[(self.df_titanic['SibSp'] == 0) & (self.df_titanic['Parch']==0)]
        wu_cuihuo = wu_peoples['Survived'].value_counts()[1]
        wu_sum = wu_peoples.count()[0]
        wu_rate = wu_cuihuo/wu_sum

        you_sum = self.sum_ - wu_sum
        you_cuihuo = self.df_titanic['Survived'].value_counts()[1] - wu_cuihuo
        you_rate = you_cuihuo/you_sum

        # return [you_cuihuo,you_sum,you_rate,'------',wu_cuihuo,wu_sum,wu_rate]

        if you_rate > wu_rate:
            return '有亲戚在船上的人的存活率{} 大于 没有亲戚在船上的存活率{}'.format(you_rate,wu_rate)
        return '没有亲戚在船上的存活率{} 大于 有亲戚在船上的人的存活率{}'.format(wu_rate, you_rate)

    # 7  从那个港口是否影响存活率
    def gangkou_survived(self):
        s_gang = self.df_titanic[self.df_titanic['Embarked'] == 'S']
        q_gang = self.df_titanic[self.df_titanic['Embarked'] == 'Q']
        c_gang = self.df_titanic[self.df_titanic['Embarked'] == 'C']
        s_rate = self._rate_survival(s_gang)
        q_rate = self._rate_survival(q_gang)
        c_rate = self._rate_survival(c_gang)
        return {'s港口':s_rate,'q港口':q_rate,'c港口':c_rate}
    # 8  不同年龄段女性的获救率
    def nv_different_age(self):
        max_survival_age = 0
        max_rate = 0
        data = self.df_titanic[self.df_titanic['Sex'].str.contains('female')]
        max_age = data['Age'].max()
        n = int(max_age//10 +1)
        # return n
        for i in range(n):
            ages = data[(i * 10 < data['Age']) & (data['Age'] <= (i + 1) * 10)]
            sum_ages = ages.count()[0]
            sum_survival = ages['Survived'].value_counts()[1]
            rate = sum_survival / sum_ages
            if rate > max_survival_age:
                max_survival_age = i
                max_rate = rate
        return '{}岁--{}岁的女人存活率最大，区间存活率为{}'.format(max_survival_age * 10, (max_survival_age + 1) * 10, max_rate)



if __name__ == '__main__':
    titanic_cc = Titanic()
    print(titanic_cc.survival_rate())
    print(titanic_cc.max_survival_age_bracket())
    print(titanic_cc.sex_survival())
    print(titanic_cc.rich_gap())
    print(titanic_cc.pclass_survivesd())
    print(titanic_cc.qinshu_survived_ziji())
    print(titanic_cc.gangkou_survived())
    print(titanic_cc.nv_different_age())

需要先下载Titanic.csv 文件，然后保存到目录中，因为是浅分析，只能固定的分析内容，烦请谅解。谢谢