I have a dataframe in R dat that contains the columns Gene, Expression, and SampleID. What I've done is sort dat such that it's grouped by Gene and arranged by descending Expression for each gene, using the following:
dat_sorted <- dat %>% select(Gene, Expression, SampleID) %>%
group_by(Gene) %>%
arrange(Gene, desc(Expression))
What I now wish to do is add a Rank column to dat_sorted which would apply a rank within each Gene group based on the Expression value, such that, for a given gene, a given sample will have a higher rank if it's expression is higher.
Here's an example of what the outcome should look like:
Gene Expression Sample Rank
ENSG00000000003 2.81561500 HSB671 1
ENSG00000000003 2.79336700 HSB431 2
ENSG00000000003 2.40009100 HSB618 3
ENSG00000000938 1.75148448 HSB671 1
ENSG00000000938 1.52182467 HSB670 2
ENSG00000000938 0.83478860 HSB414 3
ENSG00000000938 0.62174432 HSB459 4
Looks like people forgot about your question. Hope this doesn't come too late ^^
library(dplyr)
df %>% group_by(Gene) %>% mutate(Rank = dense_rank(desc(Expression)))
> df
# A tibble: 7 x 4
# Groups: Gene [2]
Gene Expression Sample Rank
<chr> <dbl> <chr> <dbl>
1 ENSG00000000003 2.82 HSB671 1
2 ENSG00000000003 2.79 HSB431 2
3 ENSG00000000003 2.40 HSB618 3
4 ENSG00000000938 1.75 HSB671 1
5 ENSG00000000938 1.52 HSB670 2
6 ENSG00000000938 0.835 HSB414 3
7 ENSG00000000938 0.622 HSB459 4
Or with base R:
df$Rank <- ave(-df$Expression, df$Gene, FUN = rank)