[
https://issues.apache.org/jira/browse/MAHOUT-1853?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15431100#comment-15431100
]
ASF GitHub Bot commented on MAHOUT-1853:
----------------------------------------
Github user dlyubimov commented on a diff in the pull request:
https://github.com/apache/mahout/pull/251#discussion_r75708308
--- Diff:
spark/src/test/scala/org/apache/mahout/cf/SimilarityAnalysisSuite.scala ---
@@ -191,14 +193,115 @@ class SimilarityAnalysisSuite extends FunSuite with
MahoutSuite with Distributed
//cross similarity
val matrixCrossCooc = drmCooc(1).checkpoint().collect
- val diff2Matrix = matrixCrossCooc.minus(matrixLLRCoocBtANonSymmetric)
+ val diff2Matrix = matrixCrossCooc.minus(matrixLLRCoocAtBNonSymmetric)
n = (new MatrixOps(m = diff2Matrix)).norm
//cooccurrence without LLR is just a A'B
//val inCoreAtB = a.transpose().times(b)
//val bp = 0
}
+ test("Cross-occurrence two IndexedDatasets"){
+ val a = dense(
+ (1, 1, 0, 0, 0),
+ (0, 0, 1, 1, 0),
+ (0, 0, 0, 0, 1),
+ (1, 0, 0, 1, 0))
+
+ val b = dense(
+ (0, 1, 1, 0),
+ (1, 1, 1, 0),
+ (0, 0, 1, 0),
+ (1, 1, 0, 1))
+
+ val users = Seq("u1", "u2", "u3", "u4")
+ val itemsA = Seq("a1", "a2", "a3", "a4", "a5")
+ val itemsB = Seq("b1", "b2", "b3", "b4")
+ val userDict = new BiDictionary(users)
+ val itemsADict = new BiDictionary(itemsA)
+ val itemsBDict = new BiDictionary(itemsB)
+
+ // this is downsampled to the top 2 values per row to match the calc
+ val matrixLLRCoocAtBNonSymmetric = dense(
+ (0.0, 1.7260924347106847, 1.7260924347106847, 0.0),
--- End diff --
this is our accepted convention, good
> Improvements to CCO (Correlated Cross-Occurrence)
> -------------------------------------------------
>
> Key: MAHOUT-1853
> URL: https://issues.apache.org/jira/browse/MAHOUT-1853
> Project: Mahout
> Issue Type: New Feature
> Affects Versions: 0.12.0
> Reporter: Andrew Palumbo
> Assignee: Pat Ferrel
> Fix For: 0.13.0
>
>
> Improvements to CCO (Correlated Cross-Occurrence) to include auto-threshold
> calculation for LLR downsampling, and possible multiple fixed thresholds for
> A’A, A’B etc. This is to account for the vast difference in dimensionality
> between indicator types.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
