MB Rule Tool - There were no rules or itemsets found that matched the given criteria

Hello Community,

unfortunately I have not found anything about this. I am using the Superstore dataset and would like to use the MB Rule tool to define the rules to perform a Market Basket analysis.

It shows that there are no rules - How does this happen? I would have thought that the rules are freely definable, only the values for Support, Lift and Confidence are lower.

Can someone please explain this to me?

Many thanks in advance

MB_Rule.yxzp

Predictive Analysis

Accepted answers

apathetichell

A few things:

1) For R tools - don't send in VW Strings. Use Strings or V Strings.

2) The minimum confidence which Alteryx sets is 0.050 - the Minimum level of support is 0.002 - you can change these inside of the MB Rules macro if you open the tool as a macro.

Once you modify the MB Rules tool you can run with Order ID/Product ID and your workflow runs fine... See the attached. There's slightly more info on this thread: https://community.alteryx.com/t5/Alteryx-Designer-Discussions/MB-rules-Error-due-to-minimum-value-of-parameters/td-p/796231

MB_Rule.yxzp

All comments

TheOC

hey @Robinvm

I'll start off by saying I'm far from the most qualified in MB - however i did have a look through @SusanCS 's post here: https://community.alteryx.com/t5/Data-Science/Market-Basket-Analysis-102-Alteryx-Designer-Python/ba-p/662029

The error is suggestive to me that it cannot find rules within your dataset, which would typically be caused by too small a dataset, or small numbers of different products within the data.

I took a look at your data, and compared it to the data used in the example post. For reference, your data contains 10,000 records, with a max count of a product being 29:

(i summarized the data, grouping by product ID, and counting number of rows).

It's worth noting that in the blog post linked, 541,000 rows are used, with a max count of product being 2369:

Are you able to try with a larger sample of your dataset?

Cheers,
TheOC

apathetichell

A few things:

1) For R tools - don't send in VW Strings. Use Strings or V Strings.

2) The minimum confidence which Alteryx sets is 0.050 - the Minimum level of support is 0.002 - you can change these inside of the MB Rules macro if you open the tool as a macro.

MB_Rule.yxzp

Robinvm

Hey @TheOC ,

thank you very much for your answer.

I know @SusanCS 's blog and I actually find it very good and helpful. I also noticed directly that Susan's data set was much larger.

However, I thought it unlikely that you shouldn't be able to apply this tool to smaller datasets. After all, there are smaller stores that sell more expensive things and therefore have fewer transactions.

I tried it once with a data set consisting of 5 rows, and there the MB-Rule tool seems to work:

Neither a warning nor an error is displayed to me.

Which may also be another interesting observation:

If I try to apply the MB Rule Tool only to the first 100 rows of the data set, then it works.

TheOC

Thanks for sharing knowledge @Robinvm and @apathetichell, I always cool to learn more about these tools.

Quick Links

Unanswered
Community Events
Groups

This months top contributors

atcodedog05 19598

Qiu 15875

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703