Fitted Values chart - how to get values?

[Dima1]

Hi All

Could you pls advise how to get fitted values based on prediction of arima/ets model for input data? I dont need prediction, but only fitted using model. Thanks!

[NeilR]

The group names are available to the R tool, as well:

AlteryxMessage(as.character(the.model$Name[1]))

[NeilR]

If you're struggling with data frames (as I do from time to time), this article might help: Code Friendly Data Structures: Data Frames in R

[cwkoops]

Thanks Neil.  That is a useful resource.  With an assist from a colleague, we got it figured out!

[NeilR]

Thanks for sharing your final solution back to the community!

[LinzHolroyd]

Thank you for sharing your solution. 

Is there a more efficient way to obtain the fitted values from the ARIMA model using the TS Model Factory macro?

The workflow worked well for 1012 records, but is taking a very long time to run for 4845 records. 

[cwkoops]

So my solution worked for 16 months, until I upgraded to Alteryx 2020.2.3.22789 (using R.3.6.3).

See attached .yxzp. It gives me the following error: 

Error:  Error in data.frame(..., check.names = FALSE) :
Info: arguments imply differing number of rows: 0, 10
Info: Calls: cbind -> cbind -> data.frame

seemingly indicating my dataframe isn't a rectangle.  

In the R tool, if I change the cbind function to rbind, it seems to work, but would really like to get it to work as it originally did without having to transpose the output.  

[amirmsaran]

@NeilR

Here is my workflow with the covariate forecast. 

[NeilR]

Hi @amirmsaran,

Unlike the ETS and ARIMA tools, the TS Covariate Forecast tool outputs the predictions from its O output anchor (rather than an R model object). Therefore you don't need to use the R tool to get these predictions, you can see them with a Browse tool...

[amirmsaran]

Thanks @NeilR but I was wondering how I could get the fitted values for 2018-2020, prior to the forecast. I wanted to have that alongside the actual data.

[NeilR]

It appears to me that the fitted values prior to the forecast are identical in the graphs for the ARIMA tool and the TS Covariate Forecast tool - it is just the forecast values that differ. So if you're after the fitted values, I would just connect the R tool to the ARIMA tool's O output anchor.