R 将 BigCartel csv 文件中的列拆分为数据框中的长格式或 data.table

Question

Big Cartel 有一个选项可以将订单导出到 csv 文件中。但是结构不太适合我需要做的分析。

这是来自 Big cartel csv 订单下载的列和行的子集（还有其他列对手头的问题不重要）。

Number, Buyer name,Items,Item count,Item total,Total price,Total shipping,Total tax,Total discount
1,jim,product_name:Plate|product_option_name:Red|quantity:1|price:9.99|total:9.99,1,9.99,11.98,1.99,0,0
2,bill,product_name:Plate|product_option_name:Green|quantity:1|price:9.99|total:9.99;product_name:Plate|product_option_name:Blue|quantity:1|price:9.99|total:9.99,2,19.98,22.98,3,0,0
3,jane,product_name:Plate|product_option_name:Red|quantity:1|price:6.99|total:6.99;product_name:Thingy|product_option_name:|quantity:1|price:9.99|total:9.99;product_name:Mug|product_option_name:Grey|quantity:1|price:10.99|total:10.99;product_name:Cup|product_option_name:Grey|quantity:1|price:9.99|total:9.99;product_name:Saucer|product_option_name:Grey|quantity:1|price:9.99|total:9.99;product_name:Stopper|product_option_name:|quantity:1|price:9.99|total:9.99,6,57.94,64.94,7,0,0
4,dale,product_name:Plate|product_option_name:Green|quantity:1|price:10.99|total:10.99,1,10.99,13.99,4.99,0,1.99

项目列可以有多个 "line-items"，以分号 (;) 作为分隔符。每个 "line-item" 有五个属性，用竖线 (|) 分隔，即 product_name、product_option_name、数量、价格和总计（即对于行）。有一列 "Item count"，它给出了 "line-items" 的数量加上（订单）总价、运费、税收和折扣的列。对于分析，我想要以下长格式的数据，其中运费、税金和折扣也被视为 'product items'。

Number Buyer name line-item    product_option_name quantity price total
1      jim        Plate        Red                 1        9.99  9.99
1      jim        shipping                         1        1.99  1.99
1      jim        tax                              0        0     0
1      jim        discount                         0        0     0
2      bill       Plate        Green               1        9.99  9.99
2      bill       Plate        Blue                1        9.99  9.99
2      bill       shipping                         1        3     3
2      bill       tax                              0        0     0
2      bill       discount                         0        0     0
3      jane       Plate        Red                 1        6.99  6.99
3      jane       Thingy                           1        9.99  9.99
3      jane       Mug          Grey                1        10.99 10.99
3      jane       Cup          Grey                1        9.99  9.99
3      jane       Saucer       Grey                1        9.99  9.99
3      jane       Stopper                          1        9.99  9.99
3      jane       shipping                         1        7     7
3      jane       tax                              0        0     0
3      jane       discount                         0        0     0
4      dale       Plate        Green               1        10.99 10.99
4      dale       shipping                         1        4.99  4.99
4      dale       tax                              0        0     
4      dale       discount                         0        -1.99 -1.99

使用 r:data.table 中的 tstrsplit() 和 r:splitstackshape 中的 cSplit() 似乎是解决方案，但我无法获得正确的语法。我也尝试了 tidyverse/dplyr 函数 separate/spread 等，但我无法得到我需要的输出。

我一直在谷歌搜索和搜索所有 SO 问题 - 有一些解决方案（这个）很接近，但 none 很适合我，因为大多数人认为范围很广'格式而不是 'long'。

Answer 1

像这样的东西可能会为您找到想要的东西。

library(dplyr)
library(tidyr)
library(stringr)

filepath <- # Path to datafile here

df <- read.csv(filepath, stringsAsFactors = FALSE)

cols <- paste0("col", 1:(max(str_count(df$Items, ";")) + 1))

df <- df %>%
      separate(col = Items, into = cols, sep = ";", fill = "right") %>%
      gather_("column", "details", cols, na.rm = TRUE) %>%
      select(-column) %>%
      separate(col = details, into = c("product_name", "product_option_name","quantity","price","total"), sep = "\|", fill = "right") %>%
      mutate(product_name = sub("^.*\:", "", product_name),
             product_option_name = sub("^.*\:", "", product_option_name),
             quantity = sub("^.*\:", "", quantity),
             price = sub("^.*\:", "", price),
             total = sub("^.*\:", "", total)) %>%
      gather("line", "item", c(Total.shipping, Total.discount, Total.tax, product_name)) %>%
      mutate(product_option_name = ifelse(line == "product_name" & product_option_name != "", product_option_name, NA),
             line_item = ifelse(line == "product_name", item, sub("^.*\.","", line)),
             price = ifelse(line == "product_name", price, item),
             price = ifelse(line_item == "discount", as.numeric(price) * (-1), price),
             quantity = ifelse(line_item %in% c("shipping","discount","tax") & price == "0", 0, quantity),
             total = as.numeric(price) * as.numeric(quantity)) %>%
      distinct() %>%
      select(Number, Buyer.name, line_item, product_option_name, quantity, price, total) %>%
      arrange(Number)

R 将 BigCartel csv 文件中的列拆分为数据框中的长格式或 data.table

R split column in BigCartel csv file into long format in dataframe or data.table

r

bigcartel

data.table

splitstackshape

tidyverse