There was a bug in this feature in the Apache PIO version that has been fixed in the SNAPSHOT. We will do a source tag to fix it before the next release . The page you reference is being changed now to advise an Apache PIO install as the root source of the project going forward. Keep and eye out here for the announcement.
On Dec 7, 2016, at 9:01 AM, Bruno LEBON <[email protected]> wrote: Hi, I am following the instructions here: http://actionml.com/docs/pio_versions <http://actionml.com/docs/pio_versions> to add the possibility to delete events that are too old to template other than Universal recommender. I want to add it to ecommerce recommendation. However it doesnt seem to work. I have the PredicitonIO 097-aml and the template ecommerce-recommendation from apache-incubator (0.4.0) I have the following 3 errors: [ERROR] [Console$] [error] /home/aml/incubator-predictionio-template-ecom-recommender/src/main/scala/DataSource.scala:21: not found: type SelfCleaningDataSource [ERROR] [Console$] [error] with SelfCleaningDataSource { [ERROR] [Console$] [error] ^ [ERROR] [Console$] [error] /home/aml/incubator-predictionio-template-ecom-recommender/src/main/scala/DataSource.scala:26: value eventWindow is not a member of org.template.ecommercerecommendation.DataSourceParams [ERROR] [Console$] [error] override def eventWindow = dsp.eventWindow [ERROR] [Console$] [error] ^ [ERROR] [Console$] [error] /home/aml/incubator-predictionio-template-ecom-recommender/src/main/scala/DataSource.scala:31: not found: value cleanPersistedPEvents [ERROR] [Console$] [error] cleanPersistedPEvents(sc) [ERROR] [Console$] [error] ^ [ERROR] [Console$] [error] three errors found The first is related to the fact that DataSourceParams has no object declared, so I added it in the declaration of the object DataSourceParams but which type should I give to it? Is it an Object? I dont know Scala, so I am not sure how to modify the code to make it work. In engine.json it refers to : "datasource": { "params" : { "name": "some-name", "appName": "autocleangites22", "eventNames": ["view"], "eventWindow": { "duration": "3650 days", "removeDuplicates": false, "compressProperties": false } } so the code accesses to eventWindow.duration. The second error says that the interface (that would be called so in Java at least) SelfCleaningDataSource is not found in PredictionIO. When I have a look at the code I find the interface, under some annotation @DeveloperApi: @DeveloperApi trait SelfCleaningDataSource { and here too (for the method generating the third error): @DeveloperApi def cleanPersistedPEvents(sc: SparkContext): Unit ={ which makes me wonder whether there is a special way to compile the source code so that the code under this annotation is to be added? Thanks in advance for your help, Bruno
