R: match values in dataframes - r

Question

I have the following list of dataframes:
a<-data.frame(
Data0=c("Y","Y","Y","Y","Y","Y","N","N","N","N","N","N"),
Data1=c(16,18,19,20,21,50,16,18,19,20,21,50),
Data2=c(2.2291,2.0743,1.9369,1.8148,1.7064,1.6102,2.2291,2.0743,1.9369,1.8148,1.7064,1.6102)
)
b<-data.frame(
Data0=c(-2 , 0 , 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ,10 ,11) ,
Data1=c(0.8891 ,0.8891,0.9051,1,0.8891,0.8891,0.7907,0.8891,0.9929,0.8891,0.8891,0.8891,0.8891)
)
dfl<-list(a,b)
I would like to make a function, for each of the dataframes in the list, that returns the numbers in the last column, for the matching variables in the previous columns.
For a, if I send "Y" and 16, I want 2.2291 and for b, if I send 5 I want 0.7907. The problem for a is that if I send a number smaller than 16, for either "Y" or "N", I want it to return the value as if it were 16, without manipulating the input variable. I manage to do this for b, but for a, where I have a "split" between "Y" and "N", I don't know how to solve this.
So far:
get_value<-function(vector){
if (length(vector)==1) {
from<-append(head(unlist(dfl[[2]][1],use.names=FALSE),-1),-9999,0)
to <-unlist(dfl[[2]][1],use.names=FALSE)
match_from<-vector > from
match_to<-vector <=to
pos<-which(match_from==match_to)
return(unlist(dfl[[2]][pos,2]))
}
else{
print("not sure what to do for a")
}
}
get_value(4)
get_value(-44)
The solution needs to take into account that sometimes for a, the columns Data0 and Data1 have switched places.
EDIT:
input/output table:
INPUT OUTPUT
c("Y",13) --> 2.2291
c("Y",50) --> 1.6102
c("N",20) --> 1.8148
c("N",50) --> 1.6102
c(-44)) --> 0.8891

Answer 1

An alternate approach using tidyverse functions could be as follows:
library(tidyverse)
library(magrittr)
get_value<-function(vector){
if (length(vector)==1)
{ df <- dfl[[2]]
k <- df %>%
arrange(Data0)%>%
filter(Data0 >= vector) %>%
select(Data1) %>%
head(1)
}
else
{df <- dfl[[1]]
k <- df %>%
arrange(Data0)%>%
filter(Data0 == vector[1]) %>%
filter(Data1 >= vector[2]) %>%
select(Data2) %>%
head(1)
}
return(k)
}
and test some outputs as:
> get_value(c(-44))
Data1
1 0.8891
> get_value(c('N',16))
Data2
1 2.2291
> get_value(c('Y',16))
Data2
1 2.2291
> get_value(c('N',12))
Data2
1 2.2291
> get_value(c('Y',11))
Data2
1 2.2291
> get_value(c('Y',18))
Data2
1 2.0743

Answer 2

Ok, I finally figured out a way, but it is not so elegant, so help would be appreciated.
get_value<-function(vector){
if (length(vector)==1) {
from<-append(head(unlist(dfl[[2]][1],use.names=FALSE),-1),-9999,0)
to <-unlist(dfl[[2]][1],use.names=FALSE)
match_from<-vector > from
match_to<-vector <= to
pos<-which(match_from==match_to)
return(unlist(dfl[[2]][pos,2]))
}
else{
pos_1<-which(unlist(dfl[[1]][1],use.names=FALSE)==vector[1])
from<-append(head(as.numeric(unlist(dfl[[1]][2],use.names=FALSE))[pos_1],-1),-9999,0)
to <-as.numeric(unlist(dfl[[1]][2],use.names=FALSE))[pos_1]
match_from<-as.numeric(vector[2]) > from
match_to <-as.numeric(vector[2]) <=to
pos<-pos_1[match_from == match_to]
return(unlist(dfl[[1]][pos,3]))
}
}