利用Python进行数据分析--美国BB Name实践（二）

将《利用Python进行数据分析》书中第2章 1880-2010年间全美婴儿姓名的项目作为练习，name数据可在GitHub中获得。
拓展练习如下：
1.计算指定名字的年度性别比例
2.计算某个名字的相对排名
3.1计算增长或减少最多的名字
3.2计算年度最流行的名字
4.分析名字趋势：长度

#导入数据
pieces=[]
years=range(1880,2011)
for year in years:
    location = 'data/names/yob%d.txt'%year
    frame = pd.read_csv(location,names=['name','sex','births'])
    frame['year']=year
    pieces.append(frame)
names = pd.concat(pieces)
#整理数据
def add_prop(group):
    group['prop']=group['births']/group['births'].sum()
    return group
names = names.groupby(['year','sex']).apply(add_prop)
#整理数据
def select_top(group):
    return group[:1000]
top1000=names.groupby(['year','sex']).apply(select_top)
top1000.index = top1000.index.droplevel().droplevel()
top1000.head()

1.计算指定名字的年度性别比例

# 计算指定名字的年度性别比例
specific = top1000[(top1000['name']=='Mary')]
specific_pivot_table = specific.pivot_table('births',index='year',columns=['sex'],aggfunc=sum)
specific_pivot_table.head()

specific_prop = specific_pivot_table.div( specific_pivot_table.sum(1),axis=0)
specific_prop.head()

specific_prop.plot(subplots=True)

2.计算指定名字的年度性别比例2

name_list= ['Harry','Mary']
name_table = top1000.pivot_table('births',columns=['name'],index=['year','sex'],aggfunc=sum)
several_name = name_table[name_list]
several_name = several_name.unstack()
several_name['Harry_F']=several_name['Harry']['F'].div(several_name['Harry'].sum(1))
several_name['Harry_M']=several_name['Harry']['M'].div(several_name['Harry'].sum(1))
several_name['Mary_F']=several_name['Mary']['F'].div(several_name['Mary'].sum(1))
several_name['Mary_M']=several_name['Mary']['M'].div(several_name['Mary'].sum(1))
several_name

several_name = several_name[['Harry_F','Harry_M','Mary_F','Mary_M']]
several_name

several_name.plot(subplots=True)

2.计算某个名字的相对排名

# 计算某个名字的相对排名
name_table = top1000.pivot_table('births',columns='year',index='name',aggfunc=sum)
name_table.head()

name_rank = name_table.rank(method='min',axis=0,ascending=False)
myname=name_rank.loc[['Mary','Jerry','Tom'],:]
myname

mynameT=myname.T
mynameT.to_csv('mynameT.csv')
mynameT

mynameT = -mynameT
mynameT

mynameT.plot(kind='line')

3.1计算增长或减少最多的名字

name_table = top1000[top1000.sex=='F'].pivot_table('births',columns='year',index='name',aggfunc=sum)
name_rank = name_table.rank(method='min',axis=0,ascending=False)
name_rankT = name_rank.T
name_rankT

rank_diff = name_rankT.diff() #求变化值
rank_diff

rank_diff_min = rank_diff.idxmin(axis=1)#名次绝对值下降最快（即名次上升最多）
rank_diff_min.to_csv('rank_diff_min.csv')
rank_diff_min
#year
#1880           NaN
#1881           Isa
#1882    Clementine
#1883       Cathryn
#1884         Belva
           ...    
#2006          Ayla
#2007         Abril
#2008         Khloe
#2009       Maliyah
#2010          Maci
#Length: 131, dtype: object

plt.figure()
rank_diff['Mary'].plot() #Mary名字变化图

3.2计算年度最流行的名字

name_rank

popular_name = name_rank.idxmin(axis=0)
popular_name.to_csv('popular_name.csv')
# 求当年最流行的名字
popular_name
#year
#1880        Mary
#1881        Mary
#1882        Mary
#1883        Mary
#1884        Mary
#          ...   
#2006       Emily
#2007       Emily
#2008        Emma
#2009    Isabella
#2010    Isabella
#Length: 131, dtype: object

popular_name.value_counts()
#Mary        76
#Jennifer    15
#Emily       12
#Jessica      9
#Lisa         8
#Linda        6
#Ashley       2
#Isabella     2
#Emma         1
#dtype: int64

4.分析名字趋势：长度

# 分析名字趋势：长度
top1000['len_of_name']=top1000['name'].str.len()#分析长度
top1000

name_len_table = top1000.pivot_table('births',index='year',columns=['len_of_name'],aggfunc='sum')
name_len_table = name_len_table.div(name_len_table.sum(1),axis=0)
name_len_table

name_len_table.plot(subplots=True,figsize=(10,10)) #子图查看不同长度的名字

name_len_table.plot(kind='line',figsize=(8,8)) #总图查看不同趋势