从R个数据框列构造JSON列 [英] Construct JSON column from R dataframe columns
本文介绍了从R个数据框列构造JSON列的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！
问题描述
这是我的数据帧：
df <- data.frame(item = c("Box 1", "Tape", "Roll 1"), Length = c(2,
10, 6), Width = c(4,3,3), Height = c(6, NA, NA), Length_units =
c("in", "ft", "yd"), option_1 = c("item_length", "item_length",
"item_length"), option_2 = c("item_width", "item_width",
"item_width"), option_3 = c("item_height", "color", NA), option_4 =
c("thickness", NA, NA), width_units = c("in", "in", "in"),
height_units = c("in", "in", NA), color = c(NA, "clear", NA),
thickness = c( "200#", NA, NA ))
item Length Width Height Length_units option_1 option_2 option_3 option_4 width_units height_units color thickness
1 Box 1 2 4 6 in item_length item_width item_height thickness in in <NA> 200#
2 Tape 10 3 NA ft item_length item_width color <NA> in in clear <NA>
3 Roll 1 6 3 NA yd item_length item_width <NA> <NA> in <NA> <NA> <NA>
{"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}
{"dimensions":"10 ft x 3 in","color":"clear"}
{"dimensions":"6 yd x 3 in"}
我非常感谢dplyr解决方案。
已编辑：我希望我的数据框如下所示
item Length Width Height Length_units option_1 option_2 option_3 option_4 width_units height_units color thickness options_json
1 Box 1 2 4 6 in item_length item_width item_height thickness in in <NA> 200# {"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}
2 Tape 10 3 NA ft item_length item_width color <NA> in in clear <NA> {"dimensions":"10 ft x 3 in","color":"clear"}
3 Roll 1 6 3 NA yd item_length item_width <NA> <NA> in <NA> <NA> <NA> {"dimensions":"6 yd x 3 in"}
推荐答案
我们可以考虑paste/unite列，然后用toJSON转换成json-由于列的大小写不同，标准化可能更好(tolower-更容易从_units列中get得到对应的列值)。循环across列(length:height)，paste(str_c)，当列中的值为非NA(使用case_when)时，相应的‘_unit’列值，则unite这些列为单列，select感兴趣的列转换为JSON(toJSON)
library(dplyr)
library(stringr)
library(jsonlite)
library(tidyr)
out <- df %>%
rename_with(tolower, everything()) %>%
mutate(across(length:height,
~ case_when(!is.na(.) ~ str_c(.x, get(str_c(cur_column(), "_units")),
sep = " ")))) %>%
unite(dimensions, length, width, height, sep = " x ", na.rm = TRUE) %>%
select(dimensions, color, thickness) %>%
toJSON(pretty = TRUE)
-输出
out
[
{
"dimensions": "2 in x 4 in x 6 in",
"thickness": "200#"
},
{
"dimensions": "10 ft x 3 in",
"color": "clear"
},
{
"dimensions": "6 yd x 3 in"
}
]
如果要创建列，请使用mutate和rowwise
df <- df %>%
rename_with(tolower, everything()) %>%
mutate(across(length:height,
~ case_when(!is.na(.) ~ str_c(.x, get(str_c(cur_column(), "_units")),
sep = " ")))) %>%
unite(dimensions, length, width, height, sep = " x ", na.rm = TRUE) %>%
select(dimensions, color, thickness) %>%
rowwise %>%
transmute(options_json = toJSON(cur_data())) %>%
ungroup %>%
bind_cols(df, .)
-输出
df
item Length Width Height Length_units option_1 option_2 option_3 option_4 width_units height_units color thickness
1 Box 1 2 4 6 in item_length item_width item_height thickness in in <NA> 200#
2 Tape 10 3 NA ft item_length item_width color <NA> in in clear <NA>
3 Roll 1 6 3 NA yd item_length item_width <NA> <NA> in <NA> <NA> <NA>
options_json
1 [{"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}]
2 [{"dimensions":"10 ft x 3 in","color":"clear"}]
3 [{"dimensions":"6 yd x 3 in"}]
json格式包括左方括号/右方括号。我们可以使用str_remove
df <- df %>%
rename_with(tolower, everything()) %>%
mutate(across(length:height,
~ case_when(!is.na(.) ~ str_c(.x, get(str_c(cur_column(), "_units")),
sep = " ")))) %>%
unite(dimensions, length, width, height, sep = " x ", na.rm = TRUE) %>%
select(dimensions, color, thickness) %>%
rowwise %>%
transmute(options_json = str_remove_all(toJSON(cur_data()), "\[|\]")) %>%
ungroup %>%
bind_cols(df, .)
-输出
item Length Width Height Length_units option_1 option_2 option_3 option_4 width_units height_units color thickness
1 Box 1 2 4 6 in item_length item_width item_height thickness in in <NA> 200#
2 Tape 10 3 NA ft item_length item_width color <NA> in in clear <NA>
3 Roll 1 6 3 NA yd item_length item_width <NA> <NA> in <NA> <NA> <NA>
options_json
1 {"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}
2 {"dimensions":"10 ft x 3 in","color":"clear"}
3 {"dimensions":"6 yd x 3 in"}
或可能unclass之后rowwise
df <- df %>%
rename_with(tolower, everything()) %>%
mutate(across(length:height,
~ case_when(!is.na(.) ~ str_c(.x, get(str_c(cur_column(), "_units")),
sep = " ")))) %>%
unite(dimensions, length, width, height, sep = " x ", na.rm = TRUE) %>%
select(dimensions, color, thickness) %>%
rowwise %>%
transmute(options_json = toJSON(keep(unclass(cur_data()),
complete.cases), auto_unbox = TRUE)) %>%
ungroup %>%
bind_cols(df, .)
-输出
> df
item Length Width Height Length_units option_1 option_2 option_3 option_4 width_units height_units color thickness
1 Box 1 2 4 6 in item_length item_width item_height thickness in in <NA> 200#
2 Tape 10 3 NA ft item_length item_width color <NA> in in clear <NA>
3 Roll 1 6 3 NA yd item_length item_width <NA> <NA> in <NA> <NA> <NA>
options_json
1 {"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}
2 {"dimensions":"10 ft x 3 in","color":"clear"}
3 {"dimensions":"6 yd x 3 in"}
> str(df)
'data.frame': 3 obs. of 14 variables:
$ item : chr "Box 1" "Tape" "Roll 1"
$ Length : num 2 10 6
$ Width : num 4 3 3
$ Height : num 6 NA NA
$ Length_units: chr "in" "ft" "yd"
$ option_1 : chr "item_length" "item_length" "item_length"
$ option_2 : chr "item_width" "item_width" "item_width"
$ option_3 : chr "item_height" "color" NA
$ option_4 : chr "thickness" NA NA
$ width_units : chr "in" "in" "in"
$ height_units: chr "in" "in" NA
$ color : chr NA "clear" NA
$ thickness : chr "200#" NA NA
$ options_json: 'json' chr "{"dimensions":"2 in x 4 in x 6 in","thickness":"200#"}" "{"dimensions":"10 ft x 3 in","color":"clear"}" "{"dimensions":"6 yd x 3 in"}"
这篇关于从R个数据框列构造JSON列的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
查看全文
