ggplot2 - R Decile percentage plotting -

i have data frame home income poll looks this:

id       income      expense 001      2389.9      1400.5 003      5499.3      2309.2 ..       ..          .. 

*this example, actual 1 has on 5k observations

i able :

1. plot decile distribution income only.
2. create variable wich asigns tenth of distribution income each home in.

1) tried not want, i'd know percentage of homes in each tenth:

> deciles<-quantile(df$income, prob = seq(0, 1, length = 11), type = 5) > deciles         0%        10%        20%        30%        40%        50%        60%      231.89    9024.48   13308.24   16945.15   21071.38   25661.58   31607.07         70%        80%        90%       100%    40360.98   52927.98   77926.47 1634433.60  

2) second part im looking this:

id       income      expense   decile 001      2389.9      1400.5    3 003      5499.3      2309.2    5 009      2245.0      1789.2    3 ..       ..          ..        .. 

thanks!

i think asking if there function inverse of quantile, scaled , ceilinged return decile number (1-10) each observation in distribution. use ecdf or write own. mine looks this:

# using convention, decile 1 highest value.  swap -x x if want change get_decile <- function(x) ceiling(10*rank(-x, ties.method="random") / length(x)) 

and plot mean income decile like:

# reproducible example! your_df <- data.frame(id=1:1e3,                       income=rnorm(1e3,5e4,2e4),                        expense=rnorm(1e3, 3e4, 1e4))  your_df$income_decile <- get_decile(your_df$income)  library(ggplot2) ggplot(your_df, aes(x=income_decile, y=income)) +      stat_summary(fun.y=mean, geom="line") +      scale_x_reverse(breaks=1:10)