数据透视

Julia 中的数据操作

Katerina Zahradova

Instructor

数据透视表

汇总数据的一种方式
重组行与列，使数据更易读

使用 unstack() 透视

unstack 结构

用 unstack() 透视

# 使用 unstack 进行透视
# unstack(DataFrame, :col_as_rows, :col_as_cols, :values, combine = f_to_aggregate)

unstack(penguins, :species, :sex, :body_mass_g, combine = median)

3×3 DataFrame
Row  species   MALE     FEMALE
     String15  Float64  Float64
_____________________________________
1    Adelie    4000.0   3400.0
2    Chinstrap 3950.0   3550.0
3    Gentoo    5500.0   4700.0

多列作为行进行透视

# 以更多列作为行进行透视
unstack(penguins, [:species, :island], :sex, :body_mass_g, combine = median)

5×4 DataFrame
Row  species    island     MALE       FEMALE
     String15   String15   Float64?   Float64?
_____________________________________________________
1    Adelie     Torgersen  4000.0    3400.0
2    Adelie     Biscoe     4000.0    3375.0
3    Adelie     Dream      3987.5    3400.0
4    Chinstrap  Dream      3950.0    3550.0
5    Gentoo     Biscoe     5500.0    4700.0

在其他位置使用多列进行透视

# 将多列作为列
unstack(penguins, :sex, [:species, :island], :body_mass_g, combine = sum)

MethodError: no method matching 
unstack(::DataFrame, ::Symbol, ::Vector{Symbol}, ::Symbol; combine = sum)

缺失值

# 某些组合存在缺失值
unstack(penguins_missing, :species, :sex, :body_mass_g, combine = median)

3×3 DataFrame
Row  species   MALE     FEMALE
     String15  Float64  Float64?
_____________________________________
1    Adelie    4000.0   3400.0
2    Chinstrap 3950.0   3550.0
3    Gentoo    5500.0   missing

替换缺失值

# 某些组合存在缺失值
unstack(penguins_missing, :species, :sex, :body_mass_g, combine = median, fill = -1)

3×3 DataFrame
Row  species   MALE     FEMALE
     String15  Float64  Float64
_____________________________________
1    Adelie    4000.0    3400.0
2    Chinstrap 3950.0    3550.0
3    Gentoo    5500.0    -1

透视表是 DataFrame

# 保存数据透视表
pivot_penguins = unstack(penguins_missing, :species, :sex, :body_mass_g, combine = median)

# 仅选择雌性企鹅
select(pivot_penguins, :species, :FEMALE)

3×2 DataFrame
Row  species    FEMALE
     String15   Float64?
_______________________________
1    Adelie      3400.0
...

让我们来练习！

Julia 中的数据操作