|
@@ -489,6 +489,17 @@ object makedata_31_bucketDataPrint_20240821 {
|
|
|
key + ":" + value
|
|
|
}.filter(_.nonEmpty).toList
|
|
|
|
|
|
+ val offlineFeatureV2 = offlineFeature.map {
|
|
|
+ case (key, value) => {
|
|
|
+ val b9FeatureSet = Set("b9_1h_ctr", "b9_1h_ctcvr", "b9_1h_cvr", "b9_1h_conver", "b9_1h_click", "b9_1h_conver*log(view)", "b9_1h_conver*ctcvr", "b9_2h_ctr", "b9_2h_ctcvr", "b9_2h_cvr", "b9_2h_conver", "b9_2h_click", "b9_2h_conver*log(view)", "b9_2h_conver*ctcvr", "b9_3h_ctr", "b9_3h_ctcvr", "b9_3h_cvr", "b9_3h_conver", "b9_3h_click", "b9_3h_conver*log(view)", "b9_3h_conver*ctcvr", "b9_6h_ctr", "b9_6h_ctcvr", "b9_6h_cvr", "b9_6h_conver", "b9_6h_click", "b9_6h_conver*log(view)", "b9_6h_conver*ctcvr", "b9_12h_ctr", "b9_12h_ctcvr", "b9_12h_cvr", "b9_12h_conver", "b9_12h_click", "b9_12h_conver*log(view)", "b9_12h_conver*ctcvr", "b9_1d_ctr", "b9_1d_ctcvr", "b9_1d_cvr", "b9_1d_conver", "b9_1d_click", "b9_1d_conver*log(view)", "b9_1d_conver*ctcvr", "b9_3d_ctr", "b9_3d_ctcvr", "b9_3d_cvr", "b9_3d_conver", "b9_3d_click", "b9_3d_conver*log(view)", "b9_3d_conver*ctcvr", "b9_7d_ctr", "b9_7d_ctcvr", "b9_7d_cvr", "b9_7d_conver", "b9_7d_click", "b9_7d_conver*log(view)", "b9_7d_conver*ctcvr", "b9_yesterday_ctr", "b9_yesterday_ctcvr", "b9_yesterday_cvr", "b9_yesterday_conver", "b9_yesterday_click", "b9_yesterday_conver*log(view)", "b9_yesterday_conver*ctcvr", "b9_today_ctr", "b9_today_ctcvr", "b9_today_cvr", "b9_today_conver", "b9_today_click", "b9_today_conver*log(view)", "b9_today_conver*ctcvr")
|
|
|
+ if (b9FeatureSet.contains(key)) {
|
|
|
+ ""
|
|
|
+ } else {
|
|
|
+ key + ":" + value
|
|
|
+ }
|
|
|
+ }
|
|
|
+ }.filter(_.nonEmpty)
|
|
|
+
|
|
|
val allFeatureV1 = allFeatureMap.map {
|
|
|
case (key, value) =>
|
|
|
key + ":" + value
|
|
@@ -504,27 +515,6 @@ object makedata_31_bucketDataPrint_20240821 {
|
|
|
}
|
|
|
}.filter(_.nonEmpty).toList
|
|
|
|
|
|
- val cidKey = "cid_" + cid + ":0.1"
|
|
|
- val allFeatureV3 = contentList.map {
|
|
|
- case (name) =>
|
|
|
- val b9FeatureSet = Set("b9_1h_ctr", "b9_1h_ctcvr", "b9_1h_cvr", "b9_1h_conver", "b9_1h_click", "b9_1h_conver*log(view)", "b9_1h_conver*ctcvr", "b9_2h_ctr", "b9_2h_ctcvr", "b9_2h_cvr", "b9_2h_conver", "b9_2h_click", "b9_2h_conver*log(view)", "b9_2h_conver*ctcvr", "b9_3h_ctr", "b9_3h_ctcvr", "b9_3h_cvr", "b9_3h_conver", "b9_3h_click", "b9_3h_conver*log(view)", "b9_3h_conver*ctcvr", "b9_6h_ctr", "b9_6h_ctcvr", "b9_6h_cvr", "b9_6h_conver", "b9_6h_click", "b9_6h_conver*log(view)", "b9_6h_conver*ctcvr", "b9_12h_ctr", "b9_12h_ctcvr", "b9_12h_cvr", "b9_12h_conver", "b9_12h_click", "b9_12h_conver*log(view)", "b9_12h_conver*ctcvr", "b9_1d_ctr", "b9_1d_ctcvr", "b9_1d_cvr", "b9_1d_conver", "b9_1d_click", "b9_1d_conver*log(view)", "b9_1d_conver*ctcvr", "b9_3d_ctr", "b9_3d_ctcvr", "b9_3d_cvr", "b9_3d_conver", "b9_3d_click", "b9_3d_conver*log(view)", "b9_3d_conver*ctcvr", "b9_7d_ctr", "b9_7d_ctcvr", "b9_7d_cvr", "b9_7d_conver", "b9_7d_click", "b9_7d_conver*log(view)", "b9_7d_conver*ctcvr", "b9_yesterday_ctr", "b9_yesterday_ctcvr", "b9_yesterday_cvr", "b9_yesterday_conver", "b9_yesterday_click", "b9_yesterday_conver*log(view)", "b9_yesterday_conver*ctcvr", "b9_today_ctr", "b9_today_ctcvr", "b9_today_cvr", "b9_today_conver", "b9_today_click", "b9_today_conver*log(view)", "b9_today_conver*ctcvr")
|
|
|
- if (b9FeatureSet.contains(name)) {
|
|
|
- if (offlineFeature.contains(name)) {
|
|
|
- name + ":" + offlineFeature(name)
|
|
|
- } else if (allFeatureMap.contains(name)) {
|
|
|
- name + ":" + allFeatureMap(name)
|
|
|
- } else {
|
|
|
- ""
|
|
|
- }
|
|
|
- } else {
|
|
|
- if (allFeatureMap.contains(name)) {
|
|
|
- name + ":" + allFeatureMap(name)
|
|
|
- } else {
|
|
|
- ""
|
|
|
- }
|
|
|
- }
|
|
|
- }.filter(_.nonEmpty) :+ cidKey
|
|
|
-
|
|
|
val ctcvrFeature = offlineFeature.map {
|
|
|
case (key, value) =>
|
|
|
if (key.contains("ctcvr") || key.contains("Ctcvr")) {
|
|
@@ -534,7 +524,7 @@ object makedata_31_bucketDataPrint_20240821 {
|
|
|
}
|
|
|
}.filter(_.nonEmpty).toList
|
|
|
|
|
|
- result.add((label, offlineFeatureList, allFeatureV1, allFeatureV2, ctcvrFeature, allFeatureV3))
|
|
|
+ result.add((label, offlineFeatureList, allFeatureV1, allFeatureV2, ctcvrFeature, offlineFeatureV2))
|
|
|
})
|
|
|
|
|
|
result.iterator
|
|
@@ -576,13 +566,13 @@ object makedata_31_bucketDataPrint_20240821 {
|
|
|
println("路径不合法,无法写入:" + ctcvrFeature)
|
|
|
}
|
|
|
|
|
|
- val allFeatureV3 = "/dw/recommend/model/33_for_check_all_v3/" + readDate
|
|
|
- if (allFeatureV3.nonEmpty && allFeatureV3.startsWith("/dw/recommend/model/")) {
|
|
|
- println("删除路径并开始数据写入:" + allFeatureV3)
|
|
|
- MyHdfsUtils.delete_hdfs_path(allFeatureV3)
|
|
|
- data2.map(r => r._1 + "\t" + r._6.mkString("\t")).saveAsTextFile(allFeatureV3, classOf[GzipCodec])
|
|
|
+ val offlineFeatureV2 = "/dw/recommend/model/33_for_check_offline_v2/" + readDate
|
|
|
+ if (offlineFeatureV2.nonEmpty && offlineFeatureV2.startsWith("/dw/recommend/model/")) {
|
|
|
+ println("删除路径并开始数据写入:" + offlineFeatureV2)
|
|
|
+ MyHdfsUtils.delete_hdfs_path(offlineFeatureV2)
|
|
|
+ data2.map(r => r._1 + "\t" + r._6.mkString("\t")).saveAsTextFile(offlineFeatureV2, classOf[GzipCodec])
|
|
|
} else {
|
|
|
- println("路径不合法,无法写入:" + allFeatureV3)
|
|
|
+ println("路径不合法,无法写入:" + offlineFeatureV2)
|
|
|
}
|
|
|
|
|
|
}
|