英文:
group by and summarise a column into defined breaks
问题
以下是翻译的代码部分：
我有以下数据：（df <- data.frame(random_str = rep("ok", 30),value = rnorm(30, mean=50, sd=25))）我想运行一些摘要统计信息，并可以这样做：df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value)# 等等......)但我还想将“value”列分成区间（例如0-33, 34-66, 67-100）。我可以手动执行此操作：df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value),# 等等......# .. 更多内容count1_33 = sum(value < 34),count34_66 = sum(value < 67 & value > 33),count67_100 = sum(value > 66))**但是否有一种方法可以自动执行这最后三个摘要，实际上我有数百万行数据，希望有数百个等距区间。**
英文:
I have the following data:
(df <- data.frame(random_str = rep("ok", 30),value = rnorm(30, mean=50, sd=25)))
I want to run a few summary statistics, and can do so:
df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value)# and so on.....)
But the value column I also want to split into breaks (e.g. 0-33, 34-66, 67-100).
I can do this manually:
df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value),# and so on.....# .. more stuff# ..count1_33 = sum(value < 34),count34_66 = sum(value < 67 & value > 33),count67_100 = sum(value > 66))
But is there a way to do these last three summaries automatically, in reality I have millions of rows of data and want 100s of equally space breaks.
答案1
得分: 1
以下是代码的翻译部分：
set.seed(100)df <- data.frame(random_str = rep("ok", 30),value = rnorm(30, mean=50, sd=25))library(dplyr)library(tidyr)df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value)) %>%bind_cols(.,df %>%group_by(random_str,range = cut(value,breaks = seq(1, by = 33,length.out = ceiling(max(value)/33) + 1))) %>%summarise( n = n(), .groups = "drop") %>%pivot_wider(-random_str, names_from = "range", values_from = "n"))#> # A tibble: 1 x 7#> random_str max min `(1,34]` `(34,67]` `(67,100]` `(100,133]`#> <chr> <dbl> <dbl> <int> <int> <int> <int>#> 1 ok 108. 21.1 5 19 5 1
更新部分的翻译：
set.seed(100)df <- data.frame(random_str = c(rep("ok", 15), rep("not", 15)),value = rnorm(30, mean=500, sd=250))library(dplyr)library(purrr)mseq <- seq(0, 900, 100)split(df, as.factor(df$random_str)) %>%lapply(., function(mdf) {map2_dbl(.x = mseq, .y = mseq + 100,.f = ~nrow(subset(mdf, value > .x & value <= .y))) %>%set_names(., map2_chr(.x = mseq, .y = mseq + 100,.f = ~paste0("(",.x, ",", .y, "]")))}) %>%bind_rows(., .id = "random_str") %>%right_join({df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value))}, .)#> Joining, by = "random_str"#> # A tibble: 2 x 13#> random_str max min `(0,100]` `(100,200]` `(200,300]` `(300,400]`#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>#> 1 not 1078. 211. 0 0 3 3#> 2 ok 722. 294. 0 0 1 2#> # ... with 6 more variables: (400,500] <dbl>, (500,600] <dbl>, (600,700] <dbl>,#> # (700,800] <dbl>, (800,900] <dbl>, (900,1000] <dbl>
希望这些翻译对你有帮助。
英文:
set.seed(100)df <- data.frame(random_str = rep("ok", 30),value = rnorm(30, mean=50, sd=25))library(dplyr)library(tidyr)df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value)) %>%bind_cols(.,df %>%group_by(random_str,range = cut(value,breaks = seq(1, by = 33,length.out = ceiling(max(value)/33) + 1))) %>%summarise( n = n(), .groups = "drop") %>%pivot_wider(-random_str, names_from = "range", values_from = "n"))#> # A tibble: 1 x 7#> random_str max min `(1,34]` `(34,67]` `(67,100]` `(100,133]`#> <chr> <dbl> <dbl> <int> <int> <int> <int>#> 1 ok 108. 21.1 5 19 5 1
Update:
I am not sure if we can get cut to include empty ranges. For this, I used purrr::map_2 to provide the ranges "manually" and set their names (looped over each random_str). Then, I bind_rows the resulted dataframes (for each group), and finally will join it with summary stats.
set.seed(100)df <- data.frame(random_str = c(rep("ok", 15), rep("not", 15)),value = rnorm(30, mean=500, sd=250))library(dplyr)library(purrr)mseq <- seq(0, 900, 100)split(df, as.factor(df$random_str)) %>%lapply(., function(mdf) {map2_dbl(.x = mseq, .y = mseq + 100,.f = ~nrow(subset(mdf, value > .x & value <= .y))) %>%set_names(., map2_chr(.x = mseq, .y = mseq + 100,.f = ~paste0("(",.x, ",", .y, "]")))}) %>%bind_rows(., .id = "random_str") %>%right_join({df %>%group_by(random_str) %>%summarise(max = max(value),min = min(value))}, .)#> Joining, by = "random_str"#> # A tibble: 2 x 13#> random_str max min `(0,100]` `(100,200]` `(200,300]` `(300,400]`#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>#> 1 not 1078. 211. 0 0 3 3#> 2 ok 722. 294. 0 0 1 2#> # ... with 6 more variables: (400,500] <dbl>, (500,600] <dbl>, (600,700] <dbl>,#> # (700,800] <dbl>, (800,900] <dbl>, (900,1000] <dbl>
<sup>Created on 2023-04-07 by the reprex package (v2.0.1)</sup>
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论