[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17004768#comment-17004768 ] Makoto Yui commented on HIVEMALL-280: - [~dharmikt] Do you have any progress? If there are something I can help, let me know. > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Assignee: Dharmik Thakkar >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16983854#comment-16983854 ] Makoto Yui commented on HIVEMALL-280: - It's just 2 pair computation but Apriori-algorithm is known to be efficient for computing k-pair rules where k >= 2. [https://medium.com/edureka/apriori-algorithm-d7cc648d4f1e] [https://www.geeksforgeeks.org/apriori-algorithm/] [https://codereview.stackexchange.com/questions/104637/apriori-algorithm-for-frequent-itemset-generation-in-java] [https://github.com/seratch/apriori4j] R's arules library support lift value as well. [http://r-statistics.co/Association-Mining-With-R.html#Caveat%20with%20using%20Lift] [https://www.kdnuggets.com/2016/04/association-rules-apriori-algorithm-tutorial.html/2] > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Assignee: Makoto Yui >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16983838#comment-16983838 ] Makoto Yui commented on HIVEMALL-280: - [~dharmikt] "A, B, A_count, B_count, AB_count, tx_total" is required to compute support/confidence/lift. Thus, I think "product 1, product 2, number of tx" is not enough. > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Assignee: Makoto Yui >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16983325#comment-16983325 ] Dharmik Thakkar commented on HIVEMALL-280: -- [~myui] Thanks! I need one clarification on the input table 'transaction'. Does the 'transaction' table look like below? !https://miro.medium.com/max/678/1*K2f-Rja-1xrqe9IdUy5Dgw.png! And in the query the arguments to lift/confidence/support would be all the three columns? Is this understanding correct? Please suggest. Thanks! > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Assignee: Makoto Yui >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16979532#comment-16979532 ] Makoto Yui commented on HIVEMALL-280: - [~dharmikt] Yes, it would be appreciated. > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649 > ] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (HIVEMALL-280) Support lift/confidence/support UDF
[ https://issues.apache.org/jira/browse/HIVEMALL-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16979429#comment-16979429 ] Dharmik Thakkar commented on HIVEMALL-280: -- [~myui] Can I work on this? > Support lift/confidence/support UDF > --- > > Key: HIVEMALL-280 > URL: https://issues.apache.org/jira/browse/HIVEMALL-280 > Project: Hivemall > Issue Type: New Feature >Reporter: Makoto Yui >Priority: Minor > > Support lift/confidence/support UDAF > [https://en.wikipedia.org/wiki/Lift_(data_mining]) > [https://towardsdatascience.com/a-gentle-introduction-on-market-basket-analysis-association-rules-fa4b986a40ce] > [https://medium.com/@samratjain/explained-market-basket-analysis-using-sql-a7434f30e649 > ] > {code:java} > select > item, other_item, > lift(...) as lift, > confidence () as confidence > from > transaction > group by > 1, 2{code} -- This message was sent by Atlassian Jira (v8.3.4#803005)