Hi all and thank you for helping me figure this out.
I would like to figure out the most common combination of products in our transactions for time buyer with in last 3 months (Jan '16 - Mar '16).
What I have done so far:
1. created a file with the first time buyers within the above mentioned time frame that includes customer # and order # and date of transaction
2. created a file with the first time buyer transactions including order #, customer #, and style/ product #
3. joined based on customer #
I still need to filter out customers who have only one proudct in the cart/transaction (trying to get to this)
My issue:
Not knowing what to do next
Not knowing if there is a way I can get to answer the question: What % or # of customers had certain common products in their cart? or the type of products that were common in transactions
If the question is not clear, please let me know because it makes a lot of sense to me in my head.
Again, thank you for your help.
Solved! Go to Solution.
Hi akaur,
It sounds like the MB Affinity tool is what you are looking for. It computes a variety of 'similarity' measures (cosine similarity/ cosinde distance/ number of co-occurences) for combinations of products.
The tool itself is in the product. We use the new Predictive District to build off of existing functionality with new tools and samples.
You can find an MB affinity sample here
I hope this helps. Let me know if you run into any additional troubles.
Dylan,
Thanks for the response. The links you sent aren't working. These maybe old links which got moved or somehting but I'm going to go throught the gallery and see if one of your suggestions can help me move toward my goal. Do you know if there are other links?
And of course, I will let you know if I just can't resolve this.
Sorry about that. I updated the links in the other post. Let me know if that works better.
Dylan,
I was able to do a search and find the appropriate post. Thanks for the updated links though.
I'm working on the solution so will let you know what happens.
Thanks...
I'm getting this error
Error: MB Affinity (9): MB Affinity: The R.exe exit code (4294967295) indicated an error.
I have order Ids and styles columns and i ran the cosine similarity affinity measure
Any ideas why?
Do you still get output results? If so, you should be able to ignore the error:
http://community.alteryx.com/t5/Advanced-Analytics/Boosted-Model-Error/td-p/5509
I didn't get any data. I only had 133k records in.
The value of row i column j is the number of times item i and item j occur together in the same transaction (i.e. B and A occur once in the same transaction, C and A never occur together, and J and P occur together in 7 transactions).