Netlogo variables as txt (csv) files - variables

Question

I am very new to Netlogo.
Linking a GIS file, I want to make a model simulating people's spending on food in reaction to the air pollution such as PM (particular matter). Of course, I also need to consider other variables such as air temperature, precipitation, and so forth. I have a panel data of food-consumption spending, air pollution and all other variables, collected by neighborhood.
My question is whether we can make a variable that changes by time, meaning when tick advances, the value of the variable changes. For example, agent may react to the air pollution in their neighborhood, and alter their food consumption behavior. I want to make a variable of "air pollution" for the neighborhood (patch) and that would change by time (following the data I collected). So somehow, I would like to link a text file of each variable and let net logo read it through when tick advances.
Is there any way to do that? Your advice would be very much appreciated.

Answer 1

I don't know how to link to a file and just pull one item. That sounds like accessing a database and I don't see or know of ( yet, just me ) a direct interface from netlogo to a database system.
Someone could write such a thing using Python, say, and use the interface to Python to pull the data you need. Or you could use the interface to R again with a little tweaking as certainly R has some very nice ways to read in CSV files and make them accessible on an item by item basis by the right clever statement in R.
Both of those would be fun to write. Anyone?
Lacking that, you COULD just read ALL of your data into NetLogo and populate , say, a list (indexed conceptually by tick) of lists ( of the 4 or 5 variables that change by tick ), then use a Netlogo statement to pull the right list indexed by tick ( plus or minus one, watch the zero versus one indexing. ) when you need it, and then parse that list to pull the variables out of it. That would certainly work.
I might write and post such a list of list solution this afternoon. Depends on the weather. Meanwhile, here's a related example I wrote once of reading in a CSV file and writing the variables to some patches-own variables, which is most of what you need.
I think this code works ( with a file to read of course. )
extensions [csv ]
globals [ XC YC ]
patches-own [
country
population
emissions_mass
vulnerability
jetfuel_civilian
climate_finance
]
to setup
clear-all
file-close
;; double the slashes so this works on a Windows computer
;; set-current-directory "C:\\Users\\wades\\Documents\\netlogo\\countries" ;
;; but simpler once just the let the user navigate to the file
;; file-open "countries.csv"
let mypath "???"
set-current-directory user-directory
file-open user-file
let headers file-read-line
print (word "There are the columns: " headers);
let headlist csv:from-row headers
let colcount length headlist
print (word "this many columns: " colcount)
let i 0
repeat colcount [
let determ item i headlist
print (word "column " i " is " determ)
set i ( i + 1 )
]
print "================================"
let nextline file-read-line
print ("skipping the second row of header info")
print nextline
print " "
while [ not file-at-end? ]
[
set nextline file-read-line
let mydata csv:from-row nextline
print "=============================================="
print (word "next incoming line to process is: " nextline )
print mydata;
set XC item 1 mydata
set YC item 2 mydata
print (word " data for patch (" XC ", " YC " )" )
let mytarget one-of patches with [ pxcor = XC and pycor = YC ]
ask mytarget
[
set country item 0 mydata
set population item 3 mydata
set emissions_mass item 4 mydata
set vulnerability item 5 mydata
set jetfuel_civilian item 6 mydata
set climate_finance item 7 mydata
set pcolor blue
set plabel country
]
print " that patch should now be blue"
]
file-close;
reset-ticks
end
to go
stop
tick
end

Answer 2

Here's sample code that reads time-series data from a csv file and puts it into a list-of-lists to pull later.
extensions [csv ]
globals [ list-of-lists ]
to setup
clear-all
reset-ticks
;; let max-rows-to-read 3 ;; for testing purposes
file-close
;; double the slashes so this works on a Windows computer
;; set-current-directory "C:\\Users\\wades\\Documents\\netlogo\\countries" ;
;; but simpler once just the let the user navigate to the file
;; file-open "countries.csv"
let mypath "???"
;; set-current-directory user-directory
file-open user-file
let headers file-read-line
print (word "There are the columns: " headers);
let headlist csv:from-row headers
let colcount length headlist
print (word "Found this many columns: " colcount)
let i 0
repeat colcount [
let determ item i headlist
print (word "column " i " is " determ)
set i ( i + 1 )
]
print "================================"
;; skip lines if you need to
; let nextline file-read-line
; print ("skipping the second row of header info")
; print nextline
; print " "
let rowcount 0
set list-of-lists [ ]
while [ not file-at-end? ]
[
set rowcount rowcount + 1
; if rowcount > max-rows-to-read [ print "list of lists: " show list-of-lists file-close stop]
let nextline file-read-line
let mydata csv:from-row nextline
print "=============================================="
print (word "next incoming line to process is: " nextline )
print mydata;
let onelist ( list
item 0 mydata
item 1 mydata
item 2 mydata )
set list-of-lists lput onelist list-of-lists
]
show list-of-lists
file-close;
end
to go
let indexer ticks ;; maybe you want this offset from ticks by some amount.
ifelse ( indexer < length list-of-lists )
[
let datalist item indexer list-of-lists
let time item 0 datalist
let temp item 1 datalist
let smog item 2 datalist
print (word "At tick " indexer " processing this list: " datalist )
print "yielding: "
print (word "At tick " indexer " time = " time " temp= " temp " and smog=" smog "\n")
]
[
print "All done. Have a nice day!"
]
tick
end