|
@@ -137,6 +137,7 @@ object makedata_recsys_83_originData_20250317 {
|
|
|
val tablePart = param.getOrElse("tablePart", "64").toInt
|
|
|
val beginStr = param.getOrElse("beginStr", "2025031700")
|
|
|
val endStr = param.getOrElse("endStr", "2025031700")
|
|
|
+ val whatPages = param.getOrElse("whatPages", "详情后沉浸页,回流后沉浸页&内页feed,首页feed,详情页,回流页").split(",").toSet
|
|
|
val whatLabel = param.getOrElse("whatLabel", "is_return_noself")
|
|
|
val fuSampleRate = param.getOrElse("fuSampleRate", "0.03").toDouble
|
|
|
val savePath = param.getOrElse("savePath", "/dw/recommend/model/83_origin_data/")
|
|
@@ -164,11 +165,10 @@ object makedata_recsys_83_originData_20250317 {
|
|
|
})
|
|
|
|
|
|
// 2.4 加载样本数据
|
|
|
- val pageSet = Set("详情后沉浸页", "回流后沉浸页&内页feed", "首页feed", "详情页", "回流页")
|
|
|
val odpsData = DataUtils.getODPSData(sc, project, table, partition, tablePart)
|
|
|
.filter(record => {
|
|
|
val page = getStringValue(record, "page")
|
|
|
- pageSet.contains(page)
|
|
|
+ whatPages.contains(page)
|
|
|
})
|
|
|
|
|
|
// 2.5 样本重采样
|