Colour palette from low to high for every row - r

Question

I'm trying to plot a geom_tile plot for a dataset, where I need to highlight the max and min values in every row (colour palette going from green to red)
Dataset:
draft_mean trim rf_pwr
1 12.0 1.0 12253
2 12.0 0.8 12052
3 12.0 0.6 12132
4 12.0 0.4 12280
5 12.0 0.2 11731
6 12.0 0.0 11317
7 12.0 -0.2 12126
8 12.0 -0.4 12288
9 12.0 -0.6 12461
10 12.0 -0.8 12791
11 12.0 -1.0 12808
12 12.2 1.0 12346
13 12.2 0.8 12041
14 12.2 0.6 12345
15 12.2 0.4 12411
16 12.2 0.2 12810
17 12.2 0.0 12993
18 12.2 -0.2 12796
19 12.2 -0.4 12411
20 12.2 -0.6 12342
21 12.2 -0.8 12671
22 12.2 -1.0 13161
ggplot(dataset, aes(trim, draft_mean)) +
geom_tile(aes(fill=rf_pwr), color="black") +
scale_fill_gradient(low= "green", high= "red") +
scale_x_reverse() +
scale_y_reverse()
This plot (image) is taking the minimum values and plotting them as green and maximum values as red. What I need help with is that I need colour palette to go from green to red (minimum to maximum) for every row of the plot (2 rows in this plot) rather than the whole plot.
For draft_mean=12.2, rf_pwr should be colour formatted from minimum to maximum for trim values.
For every value of draft_mean, I should be able to tell the trim values with lowest and highest rf_pwr.
I can plot individual draft_mean values to check, but all draft_mean values needs to be visualized together.

Answer 1

You can create a scaled variable where min = 0 and max = 1 per group like this:
require(tidyverse)
# create toy data
set.seed(1)
df <- data.frame(
draft_mean =sort(rep(c(12,12.2),11 )),
trim=rep(sample(seq(-1,1,length.out = 11), replace = F),2),
rf_pwr = sample(11000:13000,22)
)
# create a scaled variable per unique draft_mean (min = 0 and max = 1)
df <- df %>% group_by(draft_mean) %>% mutate(rf_scl = (rf_pwr-
min(rf_pwr))/(max(rf_pwr)-min(rf_pwr)))
ggplot(df, aes(trim, draft_mean)) +
geom_tile(aes(fill=rf_scl), color="black") +
scale_fill_gradient(low= "green", high= "red") +
scale_x_reverse() +
scale_y_reverse()