tag:blogger.com,1999:blog-3998628784041768080.post5389045327113926071..comments2024-03-28T03:27:07.147+00:00Comments on VizWiz: Workout Wednesday: Market Basket AnalysisVizWizhttp://www.blogger.com/profile/11408220384219218043noreply@blogger.comBlogger8125tag:blogger.com,1999:blog-3998628784041768080.post-61537882014142770262018-01-12T20:39:45.473+00:002018-01-12T20:39:45.473+00:00I've been looking into this for longer than I ...I've been looking into this for longer than I care to admit.. So far my lines of though are along creating two separate Sets (1.Paper & Tables - 2.Copiers & Chairs) and then making a set of those two sets. The 'IN' of that set should give the count of order id's that have all four of the items. <br /><br />Thanks for your response - reading your blog has taught me so much.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-67675105376721206862018-01-12T19:07:37.688+00:002018-01-12T19:07:37.688+00:00Not sure, see if you can figure it out.Not sure, see if you can figure it out.VizWizhttps://www.blogger.com/profile/11408220384219218043noreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-12385006520201233002018-01-12T16:42:01.066+00:002018-01-12T16:42:01.066+00:00Andy,
How would you suggest going about finding ...Andy, <br /><br />How would you suggest going about finding counts for more than two subcategories? For example, how many orders were there for Paper and Tables and Copiers and Chairs? Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-2500990472801409832017-10-22T09:07:44.355+01:002017-10-22T09:07:44.355+01:00Index is a counter. It’s the same as Rank_Dense no...Index is a counter. It’s the same as Rank_Dense not Rank. VizWizhttps://www.blogger.com/profile/11408220384219218043noreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-10704751624449697162017-10-22T01:28:11.020+01:002017-10-22T01:28:11.020+01:00To get rid of the duplicates, I used INDEX()<=2...To get rid of the duplicates, I used INDEX()<=2*TopN rather than RANK(). It works, but I'm always curious about application outside of this specific example. Are there any issues with using INDEX rather than RANK?Anonymoushttps://www.blogger.com/profile/09035136987604163390noreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-3798250686249709292017-10-20T09:24:11.219+01:002017-10-20T09:24:11.219+01:00Hi Andy,
Great article, may be you can also try a...Hi Andy,<br /><br />Great article, may be you can also try all records in x-axis from one data-set and others in y-axis from other data-set. I tried it a couple of months ago and it give a great approach to data. I am unable to upload any JPEG here :( otherwise shared the snapshotAnonymoushttps://www.blogger.com/profile/17930861301234869470noreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-76530724068320084012017-10-20T05:22:00.271+01:002017-10-20T05:22:00.271+01:00Go for it! That wasn’t the challenge posed though....Go for it! That wasn’t the challenge posed though. VizWizhttps://www.blogger.com/profile/11408220384219218043noreply@blogger.comtag:blogger.com,1999:blog-3998628784041768080.post-8217014170155640952017-10-19T23:34:30.343+01:002017-10-19T23:34:30.343+01:00Nice job! Maybe put it in a dashboard with heat m...Nice job! Maybe put it in a dashboard with heat map matrix at top with highlight actions between both charts?Paulhttps://www.blogger.com/profile/17275398628618843714noreply@blogger.com