I have a long data frame that looks like this:

I want a code that creates a new data frame that groups by the Product, Price and Decision and calculates:

(Sum for decision = yes) / ((Sum for decision = no) (sum for decision = yes))

So for example, for Food with the Price 1:

39 / (234 39) = 0.1428571

In the real data set I have 6 different Products and each has prices from 0 - 99.

The new data frame should look like this:

CodePudding user response：

Here's a way using proportions:

df %>% 
  group_by(Product, Price) %>% 
  mutate(prop = proportions(Sum)) %>% 
  filter(Decision == "yes")

  Product Price Decision   Sum    prop
  <chr>   <int> <chr>    <int>   <dbl>
1 Food        1 yes         39 0.143  
2 Food        2 yes       1312 0.296  
3 Clothes     1 yes        323 0.582  
4 Clothes     3 yes          3 0.00686

CodePudding user response：

Does this work:

library(dplyr)
df %>% group_by(Product, Price) %>% 
   summarise(Decision = Sum[Decision == 'yes']/sum(Sum))
`summarise()` has grouped output by 'Product'. You can override using the `.groups` argument.
# A tibble: 4 × 3
# Groups:   Product [2]
  Product Price Decision
  <chr>   <dbl>    <dbl>
1 Clothes     1  0.582  
2 Clothes     3  0.00686
3 Food        1  0.143  
4 Food        2  0.296