如何在极坐标中创建嵌套列表列

Question

使用 group_by() 和 Nest() 我可以在 R 中创建列表列，这对于将模型拟合到多个数据集非常有用。

如何在 Python 中的极坐标中实现相同的结果，使得数据列的每个元素都是一个小数据框？

library(tidyverse)

df <- structure(list(dataset_id = c(1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L), 
    day = c(1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L), recipe = structure(c(1L, 
    1L, 1L, 1L, 2L, 2L, 2L, 2L), levels = c("A", "B"), class = "factor"), 
    cum_trials = c(1000, 2000, 1000, 2000, 1000, 2000, 1000, 
    2000), cum_events = c(644L, 1287L, 643L, 1262L, 645L, 1312L, 
    655L, 1301L), cum_rate = c(0.644, 0.643, 0.643, 0.619, 0.645, 
    0.667, 0.655, 0.646)), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -8L))

df_list <- df |> group_by(dataset_id, day) |> nest()

df_list

df_list$data[[1]]

Answer 1

import polars as pl

df = pl.DataFrame({
    'a':[1,1,2,2], 
    'b':[3,4,5,6], 
    'c':[7,8,9,0],
    'string': ["hey", "you", "kids", "go"]
})

df.with_columns(my_list_col = pl.struct('b', 'string'))

a   b   c   string  my_list_col
i64 i64 i64 str     struct[2]
1   3   7   "hey"   {3,"hey"}
1   4   8   "you"   {4,"you"}
2   5   9   "kids"  {5,"kids"}
2   6   0   "go"    {6,"go"}

df.group_by('a').agg(list_col = pl.struct(cs.integer()))

a   list_col
i64 list[struct[2]]
1   [{3,"hey"}, {4,"you"}]
2   [{5,"kids"}, {6,"go"}]