R：基于另一个数据帧中的条件的数据帧的条件行总和

Question

我有一个数据框dfcompanies，其中列出了大约 6000 家公司，其标识代码为indentifier，他们的行业industry 以及他们在 132 个不同类别中的得分(NA, 0, or 1)，例如category1-4:

> dfcompanies

    identifier  industry            category1   category2   category3   category4

1   MBWS.PA     Alcoholic Beverages 1           0           1           1
2   KAER.VI     Commercial Banks    0           1           0           1
3   TIRO.VI     Commercial Banks    1           NA          1           0
4   EVNV.VI     Power Generation    0           1           1           1
5   MANV.VI     Processed Foods     NA          0           NA          NA
6   LENV.VI     Chemical            1           NA          1           1

我还有第二个数据框，dfmatcategories，它定义了industry 中存在的所有行业；在industrymat 列中列出；他们的材料类别在categorymat，并在第三列描述类别type(strength, or concern)：

> dfmatcategories

    industrymat         categorymat type

1   Alcoholic Beverages category1   concern
2   Alcoholic Beverages category2   strength
3   Alcoholic Beverages category3   strength
4   Insurance           category100 strength
5   Insurance           category99  strength
6   Chemical            category1   concern

理想情况下，我想为dfcompanies 中的每个公司获取在categorymat 中定义的材料类别中收到的积分总和，这些材料类别在新列sum_mat 中添加了type strength到dfcompanies。

> dfcompanies

    identifier  industry            category1   category2   category3   category4   sum_mat

1   MBWS.PA     Alcoholic Beverages 1           0           1           1           1
2   KAER.VI     Commercial Banks    0           1           0           1           5
3   TIRO.VI     Commercial Banks    1           NA          1           0           3
4   EVNV.VI     Power Generation    0           1           1           1           4
5   MANV.VI     Processed Foods     NA          0           NA          NA          1
6   LENV.VI     Chemical            1           NA          1           1           0

示例：

在dfcompaniesindentifierMBWS.PA的公司属于industryAlcoholic Beverages并获得分数：1category1,01,@987654 @ 在category3 中，还有1 在category4 中。

在dfmatcategories 下industrymat 为Alcoholic Beverages 行 1-3 在列 categorymat 中仅列出 category1、category2 和 category3 作为材料类别。

但是，只有category2 和category3 属于type strength。

因此对于MBWS.PAsum_mat => 1 =0（得分category2）+ 1（得分category3）。

不幸的是，我是 R 新手，完全不知道如何解决这个问题。

有人可以帮帮我吗？

最好的问候 托马斯

Answer 1

以下基于 tidyverse 的解决方案通过旋转数据来解决问题。

library(dplyr)
library(tidyr)

dfcompanies %>%
  full_join(dfmatcategories, by = c('industry' = 'industrymat')) %>%
  filter(!identifier %in% NA) %>%
  pivot_longer(category1:category4) %>%
  mutate(sum_mat = if_else(categorymat == name & 
                           type =='strength' &
                           value == 1, 1, 0)) %>%
  pivot_wider(names_from = name, values_from = value) %>%
  group_by(identifier) %>%
  mutate(sum_mat = sum(sum_mat)) %>%
  slice(1) %>%
  ungroup() %>%
  select(colnames(dfcompanies), sum_mat)

# # A tibble: 6 x 7
#   identifier industry            category1 category2 category3 category4 sum_mat
#   <chr>      <chr>                   <dbl>     <dbl>     <dbl>     <dbl>   <dbl>
# 1 EVNV.VI    Power Generation            0        NA        NA        NA      NA
# 2 KAER.VI    Commercial Banks            0        NA         0        NA      NA
# 3 LENV.VI    Chemical                    1        NA         1         1       0
# 4 MANV.VI    Processed Foods            NA        NA        NA        NA      NA
# 5 MBWS.PA    Alcoholic Beverages         1         1         1         1       2
# 6 TIRO.VI    Commercial Banks            1        NA         1        NA      NA

数据

dfcompanies <- structure(list(identifier = c("MBWS.PA", "KAER.VI", "TIRO.VI", 
"EVNV.VI", "MANV.VI", "LENV.VI"), industry = c("Alcoholic Beverages", 
"Commercial Banks", "Commercial Banks", "Power Generation", "Processed Foods", 
"Chemical"), category1 = c(1, 0, 1, 0, NA, 1), category2 = c(1, 
1, NA, 1, 0, NA), category3 = c(1, 0, 1, 1, NA, 1), category4 = c(1, 
1, 0, 1, NA, 1)), row.names = c(NA, -6L), class = c("tbl_df", 
"tbl", "data.frame"))

dfmatcategories <- structure(list(industrymat = c("Alcoholic Beverages", "Alcoholic Beverages", 
"Alcoholic Beverages", "Insurance", "Insurance", "Chemical"), 
    categorymat = c("category1", "category2", "category3", "category100", 
    "category99", "category1"), type = c("concern", "strength", 
    "strength", "strength", "strength", "concern")), row.names = c(NA, 
-6L), class = c("tbl_df", "tbl", "data.frame"))