Assigning several lists to variables whose names are contained in other variables

result <- assign(paste("subnet_",  val, sep = "")

result <- my_function(val)

-----Original Message-----
From: R-help <r-help-bounces at r-project.org> On Behalf Of Wolfgang Grond
Sent: Friday, April 9, 2021 11:49 AM
To: mailman, r-help <r-help at r-project.org>
Subject: [R] Assigning several lists to variables whose names are contained 
in
other variables

Dear all,

I'm creating a list (which is a tbl_graph) by a function, and assign the 
result to
a variable:

subnet_MYSUBNET <- my_function(MYSUBNET)

# MYSUBNET: a tbl_graph

Because there are multiple subnets to create, I can get the names of the
subnets (MYSUBNET1, MYSUBNET2, MYSUBNET3, etc.) from a row in a
dataframe column.

subnet_MYSUBNET <- my_function(datatable$column[i])

Because I know how many subnets to create - nrow(dataframe) I want to
assign the subnets to variables whose names contain the name of the subnet

For this to work I have to assign a variable name which is contained in an
other variable name:

#################

for(i in 1:nrow(datatable)) {

val <- datatable$column[i]

result <- assign(paste("subnet_",  val, sep = "")

result <- my_function(val)

}

this works in bash, but seems not to work in R - I don't succeed at least.

Am I wrong?

Where is my mistake?

Many thanks in advance for any hint.

Wolfgang



	[[alternative HTML version deleted]]

I want to assign the subnets to variables whose names contain the
name of the subnet

Dear Wolfgang,

On Fri, 9 Apr 2021 11:48:55 +0200
Wolfgang Grond <grond at numberland.de> wrote:

I want to assign the subnets to variables whose names contain the
name of the subnet

Apologies if this sounds too opinionated, but creating variable names
from variable values is a FAQ in a different dynamic language:

https://perldoc.perl.org/perlfaq7#How-can-I-use-a-variable-as-a-variable-name?

Most of the explanation doesn't apply to R, of course, but the main
idea here is to use data structures instead of causing (potential,
unlikely, but still) conflicts in the variable namespace. What if you
create a list of function values instead of just a bunch of variables?

results <- list()
for(i in 1:nrow(datatable)) {
val <- datatable$column[i]
results[[as.character(val)]] <- my_function(val)
}

Or even

results <- lapply(setNames(nm = datatable$column), my_function)

Wouldn't that be more convenient?

-- 
Best regards,
Ivan

Dear Wolfgang,

On Fri, 9 Apr 2021 11:48:55 +0200
Wolfgang Grond <grond at numberland.de> wrote:

I want to assign the subnets to variables whose names contain the
name of the subnet

Apologies if this sounds too opinionated, but creating variable names
from variable values is a FAQ in a different dynamic language:

https://perldoc.perl.org/perlfaq7#How-can-I-use-a-variable-as-a-variable-name?

Most of the explanation doesn't apply to R, of course, but the main
idea here is to use data structures instead of causing (potential,
unlikely, but still) conflicts in the variable namespace. What if you
create a list of function values instead of just a bunch of variables?

results <- list()
for(i in 1:nrow(datatable)) {
val <- datatable$column[i]
results[[as.character(val)]] <- my_function(val)
}

Or even

results <- lapply(setNames(nm = datatable$column), my_function)

Wouldn't that be more convenient?

-- 
Best regards,
Ivan