CODEWITHSUNDEEP
rdplyr
Kulluk007
Kulluk007

Reputation: 972

Computing ratios of groups in dplyr based on values from other rows

I'm trying to do some data summarization using R and dplyr. My data frame has many rows of the following form:

 color   year       score
 <fctr>  <int>       <int>

I have the same number of year rows for each of N different colors. For each of these, I have a score. Within each color (group), I'd like to compute the ratio of all of the scores to one particular year. For example:

   color   year       score
   <fctr>  <int>       <int> 
1  blue    1980        43
2  blue    1982         13
3  red     1980         330
4  red     1998         89

I'd like to augment this frame with a new column called "ratio" which is the quotient of the score of each row within each color group (e.g., blue or red) and the score of the row with a fixed year, 1980. For example:

   color   year       score    ratio
   <fctr>  <int>       <int>    
1  blue    1980        43       1
2  blue    1982         13      0.302325581
3  red     1980         330     1
4  red     1998         89      0.269696969

I know how to use mutate and summarize, but it's not clear to me how to select out the score value for a given row that meets a certain condition (in this case, the row with the year 1980 (of which we are guaranteed just one)) within each group.

What's a clean way to do this?

Upvotes: 1

Views: 2036

Answers (1)

Kulluk007
Kulluk007

Reputation: 972

akrun's comment answered my question:

mutate(ratio = score/score[year==1980])

is exactly what I needed here.

Upvotes: 4

Related Questions