Question regarding older version of Decision three tool

[Emil_Kos]

Hi,

I was looking into this Alteryx challenge and it looks like the older version of the tool has different results in the new version.

I wanted to know if anyone knows what is the difference between the 1.1 version and 1.0 version:

I checked a couple of pages for this weekly challenge but I wasn't able to find an answer. 

[ArtApa · Accepted solution]

Hi @Emil_Kos - In 2021.3 you have this opportunity to compare workflows:

You can compare both Macros and investigate every little change:

Hope that helps.

[Emil_Kos]

Hi @ArtApa,

Do you know if there was maybe a blog post what are the changes in the 1.1 version compared to 1.0?

I tried to find it.

If no one will share anything I will take a look myself but I have a feeling it might be to complicated for me to understand. 

[atcodedog05 · Accepted solution]

Hi @Emil_Kos 

As far as i know its a change in R code. And this thing has been there from a long time. I could find the below statement in 2019.2 release notes

• R has been upgraded to version 3.5.

https://help.alteryx.com/release-notes/designer/designer-20192-release-notes

Maybe 1.1 has the newer R version with that we can also assume that it might be using different packages supported in the newer version. 1.0 is still present for an older version of R for compatibility.

I would look at the R code. Hope this helps 🙂

[AkimasaKajitani · Accepted solution]

Hi @Emil_Kos 

I don't have the answer but tried to find something.

I think the big difference is using the kind of R package for predictive.

The standard rpart function of R is used in older version macro.
The AlteryxPredictive package is used in current version.

You can get the source code of it from git(https://github.com/alteryx/AlteryxPredictive).

The packages is made before 5 years ago. So I think the information is less.

[Emil_Kos]

Hi @AkimasaKajitani @atcodedog05 @ArtApa,

Thank you, everyone, for your help I will try to investigate myself in a near future. If I will do that I will post my findings here. 

[atcodedog05]

Happy to help : ) @Emil_Kos 

Interested to know about your findings.

Cheers and have a nice day!