时间戳

首页 > 解决方案 > 数据分组可与 Pandas group-by 和 grouper 相媲美

r - 数据分组可与 Pandas group-by 和 grouper 相媲美

问题描述

我有分层事件的数据集，其中一个事件有一行。

TIME               level1   level2  Occurrence
29/11/2019 00:05    A       a       1
29/11/2019 00:05    B       a       1
29/11/2019 00:07    B       b       1
29/11/2019 00:20    B       b       1
29/11/2019 00:05    B       c       1
29/11/2019 01:20    A       a       1
29/11/2019 01:25    A       a       1
29/11/2019 02:00    A       a       2
29/11/2019 02:00    B       a       1
29/11/2019 02:00    B       b       1
29/11/2019 02:35    B       b       1
29/11/2019 02:49    B       c       1

我将它与 Pandas groupby 和 grouper 聚合以获得如下输出

df_agg = df.groupby([pd.Grouper(freq='H'), 'level1', pd.Grouper('level2')])
df_agg.count()

TIME               level1   level2  Count
29/11/2019 00:00    A       a       1
                    B       a       1
                    B       b       2
                    B       c       1
29/11/2019 01:00    A       a       2
29/11/2019 02:00    A       a       2
                    B       a       1
                    B       b       2
                    B       c       1

我可以在 R 中实现类似的东西吗？

我正在附加一个命令来创建类似于我正在工作的数据集

dict = {"TIME" : ['29/11/2019  00:05:00', '29/11/2019  00:05:00', '29/11/2019  00:07:00', '29/11/2019  00:20:00',
                 '29/11/2019  00:05:00', '29/11/2019  01:20:00', '29/11/2019  01:25:00', '29/11/2019  02:00:00',
                 '29/11/2019  02:00:00', '29/11/2019  02:00:00', '29/11/2019  02:35:00', '29/11/2019  02:49:00'],
        "level1" : ["A", "B", "B", "B", "B", "A", "A", "A", "B","B", "B", "B"],
        "level2" : ["a", "a", "b", "b", "c", "a", "a", "a", "a", "b", "b","c"]}

tmp_df = pd.DataFrame(dict)
tmp_df = tmp_df.set_index('TIME')
tmp_df.index = pd.to_datetime(tmp_df.index)

标签： rdataframetime-series

解决方案

我们可以使用dplyr包：

library(dplyr)

dat %>% 
  group_by(TIME = format(dat$TIME,format='%d/%m/%Y %H:00:00'), level1, level2) %>% 
  count(name = "Count")

#> # A tibble: 9 x 4
#> # Groups:   TIME, level1, level2 [9]
#>   TIME                level1 level2 Count
#>   <chr>               <chr>  <chr>  <int>
#> 1 29/11/2019 00:00:00 A      a          1
#> 2 29/11/2019 00:00:00 B      a          1
#> 3 29/11/2019 00:00:00 B      b          2
#> 4 29/11/2019 00:00:00 B      c          1
#> 5 29/11/2019 01:00:00 A      a          2
#> 6 29/11/2019 02:00:00 A      a          1
#> 7 29/11/2019 02:00:00 B      a          1
#> 8 29/11/2019 02:00:00 B      b          2
#> 9 29/11/2019 02:00:00 B      c          1

数据：这是我使用的数据。请使用dput(dat)而不是复制/粘贴来提供您的数据。

structure(list(TIME = structure(c(1574985900, 1574985900, 1574986020, 
1574986800, 1574985900, 1574990400, 1574990700, 1574992800, 1574992800, 
1574992800, 1574994900, 1574995740), class = c("POSIXct", "POSIXt"
), tzone = "UTC"), level1 = c("A", "B", "B", "B", "B", "A", "A", 
"A", "B", "B", "B", "B"), level2 = c("a", "a", "b", "b", "c", 
"a", "a", "a", "a", "b", "b", "c"), Occurrence = c(1L, 1L, 1L, 
1L, 1L, 1L, 1L, 2L, 1L, 1L, 1L, 1L)), class = c("spec_tbl_df", 
"tbl_df", "tbl", "data.frame"), row.names = c(NA, -12L), spec = structure(list(
    cols = list(TIME = structure(list(format = "%d/%m/%Y %H:%M"), class = c("collector_datetime", 
    "collector")), level1 = structure(list(), class = c("collector_character", 
    "collector")), level2 = structure(list(), class = c("collector_character", 
    "collector")), Occurrence = structure(list(), class = c("collector_integer", 
    "collector"))), default = structure(list(), class = c("collector_guess", 
    "collector")), skip = 1), class = "col_spec"))

推荐阅读