Forum Discussion
Calling a column parameter in a custom function
Hello,
I have created a custom function in power bi that outputs a data table for an Arima model. I was able to successfully call a "dataset" parameter, so the user could choose a specific table, but when I tried to do the same with a "column" parameter in my function it did not work. Here is my code,
fnr = [
Arima.model=
(dataset as table,column as list) as table=>
let
RScript=R.Execute("
library(forecast)
library(tseries)
Values<-dataset[,4]
ValuesTs<-ts(Values,start=c(1999,1),frequency=4,end=c(2019,4))
fit<-auto.arima(ValuesTs,stepwise=FALSE,approximation=FALSE)
df<-data.frame(forecast(fit,h=16))
df['Years']<-seq(2020,2035,1)
",[dataset=dataset]),
output=RScript
in
output
],
As you can see, I am currently just grabbing the column I need by using the "dataset" parameter but I would like to replace that with my "column" parameter. I have tried many different ways but I am thinking that I simply have some formatting issues that are making this difficult. Any help would be greatly appreciated. Thanks.
2 Replies
Anonymous ,
What does the error message say? How about change the parameter with other names like "param1", "param2"?
Regards,
Jimmy Tao
The error is: object 'column' can't be found, even though I do select a column for that parameter when I invoke the function
Which is why I figured I had a formatting error when I would try to reference that parameter. This is my code when I tried a different parameter name instead of 'column', unfortunately I got the same error message:
fnr = [
Arima.model=(dataset as table,param1 as list) as table=>
let
RScript=R.Execute("
library(forecast)
library(tseries)
Values<-param1
ValuesTs<-ts(Values,start=c(1999,1),frequency=4,end=c(2019,4))
fit<-auto.arima(ValuesTs,stepwise=FALSE,approximation=FALSE)
df<-data.frame(forecast(fit,h=16))
df['Years']<-seq(2020,2035,1)",[dataset=dataset]),
output=RScript
in
output
],
