123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338 |
- package examples.extractor;
- import java.util.*;
- import java.util.stream.Collectors;
- public class RankExtractorItemFeatureV2 {
- public static Map<String, String> getItemRateFeature(Map<String, String> maps) {
- double d;
- Map<String, Double> result = new HashMap<>();
- d = ExtractorUtils.division("i_1day_exp_cnt", "i_1day_click_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_ctr",d);
- }
- d = ExtractorUtils.division("i_1day_exp_cnt", "i_1day_share_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_str",d);
- }
- d = ExtractorUtils.division("i_1day_exp_cnt", "i_1day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_rov",d);
- }
- d = ExtractorUtils.division("i_1day_share_cnt", "i_1day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_ros",d);
- }
- d = ExtractorUtils.division("i_3day_exp_cnt", "i_3day_click_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3day_ctr",d);
- }
- d = ExtractorUtils.division("i_3day_exp_cnt", "i_3day_share_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3day_str",d);
- }
- d = ExtractorUtils.division("i_3day_exp_cnt", "i_3day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3day_rov",d);
- }
- d = ExtractorUtils.division("i_3day_share_cnt", "i_3day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3day_ros",d);
- }
- d = ExtractorUtils.division("i_7day_exp_cnt", "i_7day_click_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_7day_ctr",d);
- }
- d = ExtractorUtils.division("i_7day_exp_cnt", "i_7day_share_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_7day_str",d);
- }
- d = ExtractorUtils.division("i_7day_exp_cnt", "i_7day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_7day_rov",d);
- }
- d = ExtractorUtils.division("i_7day_share_cnt", "i_7day_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_7day_ros",d);
- }
- d = ExtractorUtils.division("i_3month_exp_cnt", "i_3month_click_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3month_ctr",d);
- }
- d = ExtractorUtils.division("i_3month_exp_cnt", "i_3month_share_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3month_str",d);
- }
- d = ExtractorUtils.division("i_3month_exp_cnt", "i_3month_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3month_rov",d);
- }
- d = ExtractorUtils.division("i_3month_share_cnt", "i_3month_return_cnt", maps);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_3month_ros",d);
- }
- Map<String, String> result2 = new HashMap<>();
- for (Map.Entry<String, Double> entry : result.entrySet()){
- result2.put(entry.getKey(), String.valueOf(entry.getValue()));
- }
- return result2;
- }
- public static Map<String, String> getItemRealtimeTrend(Map<String, Map<String, Double>> maps, String date, String hour){
- Map<String, Double> result1 = new HashMap<>();
- Map<String, Double> result2 = new HashMap<>();
- if (date.isEmpty() || hour.isEmpty()){
- return rateFeatureChange(result1);
- }
- int N = 6;
- List<String> hourStrs = ExtractorUtils.generateHourStrings(date + hour, N);
- String key;
- key = "share_uv_list_1day";
- if (maps.containsKey(key)){
- Map<String, Double> fList = maps.get(key);
- List<Double> arrs = hourStrs.stream().map(r -> fList.getOrDefault(r, 0.0D)).collect(Collectors.toList());
- Collections.reverse(arrs);
- result1.put(key+"_"+N+"_avg", ExtractorUtils.calculateAverage(arrs));
- result1.put(key+"_"+N+"_var", ExtractorUtils.calculateVariance(arrs));
- List<Double> arrsDiff = ExtractorUtils.calculateDifferences(arrs);
- result2.put(key+"_diff_"+N+"_avg", ExtractorUtils.calculateAverage(arrsDiff));
- result2.put(key+"_diff_"+N+"_var", ExtractorUtils.calculateVariance(arrsDiff));
- }
- key = "return_uv_list_1day";
- if (maps.containsKey(key)){
- Map<String, Double> fList = maps.get(key);
- List<Double> arrs = hourStrs.stream().map(r -> fList.getOrDefault(r, 0.0D)).collect(Collectors.toList());
- Collections.reverse(arrs);
- result1.put(key+"_"+N+"_avg", ExtractorUtils.calculateAverage(arrs));
- result1.put(key+"_"+N+"_var", ExtractorUtils.calculateVariance(arrs));
- List<Double> arrsDiff = ExtractorUtils.calculateDifferences(arrs);
- result2.put(key+"_diff_"+N+"_avg", ExtractorUtils.calculateAverage(arrsDiff));
- result2.put(key+"_diff_"+N+"_var", ExtractorUtils.calculateVariance(arrsDiff));
- }
- key = "share_uv_list_1h";
- if (maps.containsKey(key)){
- Map<String, Double> fList = maps.get(key);
- List<Double> arrs = hourStrs.stream().map(r -> fList.getOrDefault(r, 0.0D)).collect(Collectors.toList());
- Collections.reverse(arrs);
- result1.put(key+"_"+N+"_avg", ExtractorUtils.calculateAverage(arrs));
- result1.put(key+"_"+N+"_var", ExtractorUtils.calculateVariance(arrs));
- List<Double> arrsDiff = ExtractorUtils.calculateDifferences(arrs);
- result2.put(key+"_diff_"+N+"_avg", ExtractorUtils.calculateAverage(arrsDiff));
- result2.put(key+"_diff_"+N+"_var", ExtractorUtils.calculateVariance(arrsDiff));
- }
- key = "return_uv_list_1h";
- if (maps.containsKey(key)){
- Map<String, Double> fList = maps.get(key);
- List<Double> arrs = hourStrs.stream().map(r -> fList.getOrDefault(r, 0.0D)).collect(Collectors.toList());
- Collections.reverse(arrs);
- result1.put(key+"_"+N+"_avg", ExtractorUtils.calculateAverage(arrs));
- result1.put(key+"_"+N+"_var", ExtractorUtils.calculateVariance(arrs));
- List<Double> arrsDiff = ExtractorUtils.calculateDifferences(arrs);
- result2.put(key+"_diff_"+N+"_avg", ExtractorUtils.calculateAverage(arrsDiff));
- result2.put(key+"_diff_"+N+"_var", ExtractorUtils.calculateVariance(arrsDiff));
- }
- Map<String, String> result = new HashMap<>();
- for (Map.Entry<String, Double> entry : result1.entrySet()){
- result.put(entry.getKey(), String.valueOf(entry.getValue()));
- }
- for (Map.Entry<String, Double> entry : result2.entrySet()){
- result.put(entry.getKey(), String.valueOf(entry.getValue()));
- }
- return result;
- }
- public static Map<String, String> rateFeatureChange(Map<String, Double> maps){
- Map<String, String> result = new HashMap<>();
- for (Map.Entry<String, Double> entry : maps.entrySet()){
- int value = ExtractorUtils.ceilLogRate(entry.getValue());
- result.put(entry.getKey(), String.valueOf(value));
- }
- return result;
- }
- public static Map<String, String> cntFeatureChange4Double(Map<String, Double> maps){
- Map<String, String> result = new HashMap<>();
- for (Map.Entry<String, Double> entry : maps.entrySet()){
- int value = ExtractorUtils.bucketCnt(entry.getValue());
- result.put(entry.getKey(), String.valueOf(value));
- }
- return result;
- }
- public static Map<String, String> cntFeatureChange(Map<String, String> maps,
- Set<String> names){
- Map<String, String> result = new HashMap<>();
- for (Map.Entry<String, String> entry : maps.entrySet()){
- if (!names.contains(entry.getKey())){
- continue;
- }
- int value = ExtractorUtils.bucketCnt(Double.valueOf(entry.getValue()));
- result.put(entry.getKey(), String.valueOf(value));
- }
- return result;
- }
- public static Map<String, String> getItemRealtimeCnt(Map<String, Map<String, Double>> maps,
- Set<String> names,
- String date, String hour){
- Map<String, String> result = new HashMap<>();
- if (date.isEmpty() || hour.isEmpty()){
- return result;
- }
- String dateHour = ExtractorUtils.subtractHours(date + hour, 0);
- for (Map.Entry<String, Map<String, Double>> entry : maps.entrySet()){
- if (!names.contains(entry.getKey())){
- continue;
- }
- Double num = entry.getValue().getOrDefault(dateHour, 0.0);
- if (!ExtractorUtils.isDoubleEqualToZero(num)){
- result.put(entry.getKey(), String.valueOf(num));
- }
- }
- return result;
- }
- public static Map<String, String> getItemRealtimeRate(Map<String, Map<String, Double>> maps,
- String date, String hour){
- Map<String, Double> result = new HashMap<>();
- if (date.isEmpty() || hour.isEmpty()){
- return rateFeatureChange(result);
- }
- String dateHour = ExtractorUtils.subtractHours(date + hour, 0);
- double d, d1, d2;
- String k1, k2;
- k1 = "view_pv_list_1day";
- k2 = "play_pv_list_1day";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_ctr_rt", d);
- }
- }
- k1 = "view_pv_list_1day";
- k2 = "share_pv_list_1day";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_str_rt", d);
- }
- }
- k1 = "share_pv_list_1day";
- k2 = "return_uv_list_1day";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_ros_rt", d);
- }
- }
- k1 = "view_pv_list_1day";
- k2 = "return_uv_list_1day";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1day_rov_rt", d);
- }
- }
- //---
- k1 = "view_pv_list_1h";
- k2 = "play_pv_list_1h";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1h_ctr_rt", d);
- }
- }
- k1 = "view_pv_list_1h";
- k2 = "share_pv_list_1h";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1h_str_rt", d);
- }
- }
- k1 = "share_pv_list_1day";
- k2 = "return_uv_list_1h";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1h_ros_rt", d);
- }
- }
- k1 = "view_pv_list_1h";
- k2 = "return_uv_list_1h";
- if (maps.containsKey(k1) && maps.containsKey(k2)){
- d1 = maps.get(k1).getOrDefault(dateHour, 0.0);
- d2 = maps.get(k2).getOrDefault(dateHour, 0.0);
- d = ExtractorUtils.divisionDouble(d1, d2);
- if (!ExtractorUtils.isDoubleEqualToZero(d)){
- result.put("i_1h_rov_rt", d);
- }
- }
- Map<String, String> result2 = new HashMap<>();
- for (Map.Entry<String, Double> entry : result.entrySet()){
- result2.put(entry.getKey(), String.valueOf(entry.getValue()));
- }
- return result2;
- }
- public static void main(String[] args) {
- String s1 = "share_uv_list_1day";
- String s2 = "2024011300:2,2024011301:2,2024011304:2,2024011309:3,2024011311:3,2024011314:4,2024011315:4,2024011321:1,2024011323:1,2024011400:1,2024011401:1,2024011404:1,2024011406:1,2024011407:1,2024011408:1,2024011410:1,2024011423:1,2024011302:2,2024011305:2,2024011312:4,2024011313:4,2024011317:4,2024011318:4,2024011319:3,2024011320:1,2024011403:1,2024011409:1,2024011411:1,2024011419:1,2024011420:1,2024011422:1,2024011303:2,2024011306:2,2024011307:2,2024011308:2,2024011310:3,2024011316:4,2024011322:1,2024011402:1,2024011405:1,2024011421:1";
- Map<String, Double> m1 = new HashMap<>();
- Map<String, Map<String, Double>> maps = new HashMap<>();
- for (String s : s2.split(",")){
- String s3 = s.split(":")[0];
- String s4 = s.split(":")[1];
- m1.put(s3, Double.valueOf(s4));
- }
- maps.put(s1, m1);
- String date = "20240114";
- String hour = "20";
- System.out.println(getItemRealtimeTrend(maps, date, hour));
- }
- }
|