Find the mean of one variable subseted by another variable - r

Question

I have a list of identical dataframes. Each data frame contains columns with unique variables (temp/DO) and with repeated variables (eg-t1).
[[1]]
temp DO t1
1 4 1
3 9 1
5 7 1
I want to find the mean of DO when the temperature is equal to t1.
t1 represents a specific temperature, but the value varies for each data frame in the list so I can't specify an actual value.
So far I've tried writing a function
hvod<-function(DO, temp, depth){
hDO<-DO[which(temp==t1[1])]
mHDO<-mean(hDO)
htemp<-temp[which(temp=t1[1])]
mhtemp<-mean(htemp)
}
hfit<-hvod(data$DO, data$temp, data$depth)
But for whatever reason t1 is not recognized. Any ideas on the function OR
a way to combine select (dplyr function) and lapply to solve this?
I've seen similar posts put none that apply to the issue of a specific value (t1) that changes for each data frame.

Answer 1

I would just take the dataframe as argument and do rest of the logic inside function as it gives more control to the function. Something like this would work,
hvod<-function(data){
temp <- data$temp
t1 <- data$t1
DO <- data$DO
hDO<-DO[which(temp==t1[1])]
mHDO<-mean(hDO)
htemp<-temp[which(temp=t1[1])]
mhtemp<-mean(htemp)
}

Answer 2

You can try using dplyr::bind_rows function to combine all data.frames from list in one data.frame.
Then group on data.frame number to find the mean of DO for rows having temp==t1 as:
library(dplyr)
bind_rows(ll, .id = "DF_Name") %>%
group_by(DF_Name) %>%
filter(temp==t1) %>%
summarise(MeanDO = mean(DO)) %>%
as.data.frame()
# DF_Name MeanDO
# 1 1 4.0
# 2 2 6.5
# 3 3 6.7
Data:
df1 <- read.table(text =
"temp DO t1
1 4 1
3 9 1
5 7 1",
header = TRUE)
df2 <- read.table(text =
"temp DO t1
3 4 3
3 9 3
5 7 1",
header = TRUE)
df3 <- read.table(text =
"temp DO t1
2 4 2
2 9 2
2 7 2",
header = TRUE)
ll <- list(df1, df2, df3)

Answer 3

Thank you Thiloshon and MKR for the help! I had initial combined the data I needed into one list of data frames but to answer this I actually had my data in separate data frames (fitsObs and df1).
The variables I was working with in the code were 1 to 1, so by finding the range where depth and d2 were the same (I used temp and t1 in the example), I could find the mean over that range .
for(i in 1:1044){
df1 <- GLNPOsurveyCTD$data[[i]]
fitObs <- fitTp2(-df1$depth, df1$temp)
deptho <- -abs(df1$depth) #defining temp and depth in the loop
to <- df1$temp
do <- df1$DO
xx <- which(deptho <= fitObs$d2) #mean over range xx
mhtemp <- mean(to[xx], na.rm=TRUE)
mHDO <- mean(do[xx], na.rm=TRUE)
}