代码优化

modules/rankingCalculator.dos

@@ -4,6 +4,39 @@ use fundit::sqlUtilities
 use fundit::dataPuller
 use fundit::dataSaver
 
+/*
+ *   汇集所有参与排名的指标信息
+ * 
+ */
+def get_indicator_info() {
+
+    ids = [1,
+           2, 6, 9, 10, 11, 12, 21, 50, 52, 59,
+           14, 15, 16, 17, 18, 19, 40, 58,
+           37, 38, 41, 42, 43, 44, 45, 46, 47, 48, 49,
+           33, 34, 35, 36,
+           66, 53, 54, 55, 56, 57
+          ];
+
+    names = ['ret',
+             'maxdrawdown', 'kurtosis', 'skewness', 'stddev', 'alpha', 'beta', 'downsidedev', 'maxdrawdown_months', 'maxdrawdown_recoverymonths', 'winrate',
+             'kapparatio', 'treynorratio', 'jensen', 'omegaratio', 'sharperatio', 'sortinoratio_MAR', 'calmarratio', 'sortinoratio',
+             'per_con', 'info_ratio', 'var', 'cvar', 'smddvar', 'smddcvar', 'smdd_lpm1', 'smdd_lpm2', 'smdd_downside_dev', 'tracking_error', 'm2',
+             'upsidecapture_ret', 'downsidecapture_ret', 'upsidecapture_ratio', 'downsidecapture_ratio',
+             'stability', 'jc_stddev', 'gzstyle_stddev', 'gzstrategy_stddev', 'zz_stddev', 'zx_stddev'
+            ];
+
+    is_ASCs = [false,
+               true, true, false, true, false, false, true, true, true, false,
+               false, false, false, false, false, false, false, false,
+               false, false, true, true, true, true, true, true, true, true, false,
+               false, false, false, true,
+               true, true, true, true, true, true
+              ];
+
+    return table(names AS name, ids AS id, is_ASCs AS is_ASC);
+
+}
 
 /*
  *   计算收益率排名
@@ -12,6 +45,9 @@ use fundit::dataSaver
  */
 def cal_ret_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
+
 	table_desc = get_performance_table_description(entity_type);
 
 	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
@@ -178,162 +214,255 @@ defg perRank(x, is_ASC) {
  *   @param indicator_name <STRING>: 指标字段名
  *   @param indicator_id <INT>：指标ID
  *   @param is_ASC <BOOL>: 是否排正序
- *   @param ranking_by <STRING>: 'strategy', 'substrategy', 'factor', 'catavg'
+ *   @param ranking_by <STRING>: 'strategy', 'substrategy', 'factor_id', 'catavg'
+ *   
+ *   TODO: portfolio, cf, manager, company, 
+ *   TODO: bfi & category
+ * 
+ */
+def gen_ranking_sql0(data_table, indicator_table, ranking_by) {
+
+	for(indicator in indicator_table) {
+
+	    // 近1月和近3月排名仅对收益有效，为了满足表结构的要求，需要建立几个”假”字段，并用NULL赋值
+	    t_tmp = table(1000:0, ['indicator_id', 'indicator_1m', 'absrank_1m', 'perrank_1m',
+	                                           'indicator_3m', 'absrank_3m', 'perrank_3m'],
+	                          [INT, DOUBLE, INT, INT, DOUBLE, INT, INT]);
+	    INSERT INTO t_tmp VALUES (indicator.id, double(NULL), int(NULL), int(NULL), double(NULL), int(NULL), int(NULL));
+	
+	    t_ranking = sql(select = (sqlCol('entity_id'), sqlCol('end_date'), sqlCol(ranking_by), sqlCol('indicator_id'),
+	                              sqlCol('indicator_1m'), sqlCol('absrank_1m'), sqlCol('perrank_1m'),
+	                              sqlCol('indicator_3m'), sqlCol('absrank_3m'), sqlCol('perrank_3m'),
+	                              // 与 MySQL 不同，这里统一把近4年和成立以来的排名去掉
+	                              sqlCol(indicator.name + '_6m',,'indicator_6m'),
+	                              sqlCol(indicator.name + '_6m', rank{, indicator.is_ASC}, 'absrank_6m'),
+	                              sqlCol(indicator.name + '_6m', perRank{, indicator.is_ASC}, 'perrank_6m'),
+	                              sqlCol(indicator.name + '_1y',,'indicator_1y'),
+	                              sqlCol(indicator.name + '_1y', rank{, indicator.is_ASC}, 'absrank_1y'),
+	                              sqlCol(indicator.name + '_1y', perRank{, indicator.is_ASC}, 'perrank_1y'),
+	                              sqlCol(indicator.name + '_2y',,'indicator_2y'),
+	                              sqlCol(indicator.name + '_2y', rank{, indicator.is_ASC}, 'absrank_2y'),
+	                              sqlCol(indicator.name + '_2y', perRank{, indicator.is_ASC}, 'perrank_2y'),
+	                              sqlCol(indicator.name + '_3y',,'indicator_3y'),
+	                              sqlCol(indicator.name + '_3y', rank{, indicator.is_ASC}, 'absrank_3y'),
+	                              sqlCol(indicator.name + '_3y', perRank{, indicator.is_ASC}, 'perrank_3y'),
+	                              sqlCol(indicator.name + '_5y',,'indicator_5y'),
+	                              sqlCol(indicator.name + '_5y', rank{, indicator.is_ASC}, 'absrank_5y'),
+	                              sqlCol(indicator.name + '_5y', perRank{, indicator.is_ASC}, 'perrank_5y'),
+	                              sqlCol(indicator.name + '_10y',,'indicator_10y'),
+	                              sqlCol(indicator.name + '_10y', rank{, indicator.is_ASC}, 'absrank_10y'),
+	                              sqlCol(indicator.name + '_10y', perRank{, indicator.is_ASC}, 'perrank_10y'),
+	                              sqlCol(indicator.name + '_ytd',,'indicator_ytd'),
+	                              sqlCol(indicator.name + '_ytd', rank{, indicator.is_ASC}, 'absrank_ytd'),
+	                              sqlCol(indicator.name + '_ytd', perRank{, indicator.is_ASC}, 'perrank_ytd')
+	                              ), 
+	                    from = cj(data_table, t_tmp),
+	                    where = <_$ranking_by IS NOT NULL>,
+	                    groupBy = (sqlCol(ranking_by), sqlCol('end_date')),
+	                    groupFlag = 0 ).eval(); // context by
+	
+	
+	    // 近1月和近3月排名仅对收益有效，为了满足表结构的要求，需要建立几个”假”字段，并用NULL赋值
+	    t_tmp = table(1000:0, ['indicator_id', 'avg_1m', 'avg_1m_cnt', 'perrank_percent_5_1m', 'perrank_percent_10_1m', 'perrank_percent_25_1m',
+	                           'perrank_percent_50_1m', 'perrank_percent_75_1m', 'perrank_percent_90_1m', 'perrank_percent_95_1m', 'best_1m', 'worst_1m',
+	                           'avg_3m', 'avg_3m_cnt', 'perrank_percent_5_3m', 'perrank_percent_10_3m', 'perrank_percent_25_3m',
+	                           'perrank_percent_50_3m', 'perrank_percent_75_3m', 'perrank_percent_90_3m', 'perrank_percent_95_3m', 'best_3m', 'worst_3m'],
+	                          [INT, DOUBLE, INT, DOUBLE, DOUBLE, DOUBLE,
+	                           DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE,
+	                           DOUBLE, INT, DOUBLE, DOUBLE, DOUBLE,
+	                           DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE]);
+	
+	    INSERT INTO t_tmp VALUES (indicator.id, double(NULL), int(NULL), double(NULL), double(NULL), double(NULL),
+	                              double(NULL), double(NULL), double(NULL), double(NULL), double(NULL), double(NULL),
+	                              double(NULL), int(NULL), double(NULL), double(NULL), double(NULL),
+	                              double(NULL), double(NULL), double(NULL), double(NULL), double(NULL), double(NULL)); 
+	
+	    t_ranking_num = sql(select = (sqlCol('end_date'), sqlCol(ranking_by), sqlCol('raise_type', mean, 'raise_type'), sqlCol('indicator_id', mean,'indicator_id'),
+	                                  sqlCol('avg_1m', mean, 'avg_1m'), sqlCol('avg_1m_cnt', mean, 'avg_1m_cnt'),
+	                                  sqlCol('perrank_percent_5_1m', mean, 'perrank_percent_5_1m'),
+	                                  sqlCol('perrank_percent_10_1m', mean, 'perrank_percent_10_1m'),
+	                                  sqlCol('perrank_percent_25_1m', mean, 'perrank_percent_25_1m'),
+	                                  sqlCol('perrank_percent_50_1m', mean, 'perrank_percent_50_1m'),
+	                                  sqlCol('perrank_percent_75_1m', mean, 'perrank_percent_75_1m'),
+	                                  sqlCol('perrank_percent_90_1m', mean, 'perrank_percent_90_1m'),
+	                                  sqlCol('perrank_percent_95_1m', mean, 'perrank_percent_95_1m'), 
+	                                  sqlCol('best_1m', mean, 'best_1m'), sqlCol('worst_1m', mean, 'worst_1m'),
+	                                  sqlCol('avg_3m', mean, 'avg_3m'), sqlCol('avg_3m_cnt', mean, 'avg_3m_cnt'),
+	                                  sqlCol('perrank_percent_5_3m', mean, 'perrank_percent_5_3m'),
+	                                  sqlCol('perrank_percent_10_3m', mean, 'perrank_percent_10_3m'),
+	                                  sqlCol('perrank_percent_25_3m', mean, 'perrank_percent_25_3m'),
+	                                  sqlCol('perrank_percent_50_3m', mean, 'perrank_percent_50_3m'),
+	                                  sqlCol('perrank_percent_75_3m', mean, 'perrank_percent_75_3m'),
+	                                  sqlCol('perrank_percent_90_3m', mean, 'perrank_percent_90_3m'),
+	                                  sqlCol('perrank_percent_95_3m', mean, 'perrank_percent_95_3m'), 
+	                                  sqlCol('best_3m', mean, 'best_3m'), sqlCol('worst_3m', mean, 'worst_3m'),
+	                                  // 与 MySQL 不同，这里统一把近4年和成立以来的排名去掉
+	                                  sqlCol(indicator.name + '_6m', mean, 'avg_6m'), sqlCol(indicator.name + '_6m', count, 'avg_6m_cnt'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_6m'),
+	                                  sqlCol(indicator.name + '_6m', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_6m'),
+	                                  sqlCol(indicator.name + '_6m', iif(indicator.is_ASC, min, max), 'best_6m'),
+	                                  sqlCol(indicator.name + '_6m', iif(indicator.is_ASC, max, min), 'worst_6m'),
+	                                  sqlCol(indicator.name + '_1y', mean, 'avg_1y'), sqlCol(indicator.name + '_1y', count, 'avg_1y_cnt'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_1y'),
+	                                  sqlCol(indicator.name + '_1y', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_1y'),
+	                                  sqlCol(indicator.name + '_1y', iif(indicator.is_ASC, min, max), 'best_1y'),
+	                                  sqlCol(indicator.name + '_1y', iif(indicator.is_ASC, max, min), 'worst_1y'),
+	                                  sqlCol(indicator.name + '_2y', mean, 'avg_2y'), sqlCol(indicator.name + '_2y', count, 'avg_2y_cnt'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_2y'),
+	                                  sqlCol(indicator.name + '_2y', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_2y'),
+	                                  sqlCol(indicator.name + '_2y', iif(indicator.is_ASC, min, max), 'best_2y'),
+	                                  sqlCol(indicator.name + '_2y', iif(indicator.is_ASC, max, min), 'worst_2y'),
+	                                  sqlCol(indicator.name + '_3y', mean, 'avg_3y'), sqlCol(indicator.name + '_3y', count, 'avg_3y_cnt'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_3y'),
+	                                  sqlCol(indicator.name + '_3y', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_3y'),
+	                                  sqlCol(indicator.name + '_3y', iif(indicator.is_ASC, min, max), 'best_3y'),
+	                                  sqlCol(indicator.name + '_3y', iif(indicator.is_ASC, max, min), 'worst_3y'),
+	                                  sqlCol(indicator.name + '_5y', mean, 'avg_5y'), sqlCol(indicator.name + '_5y', count, 'avg_5y_cnt'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_5y'),
+	                                  sqlCol(indicator.name + '_5y', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_5y'),
+	                                  sqlCol(indicator.name + '_5y', iif(indicator.is_ASC, min, max), 'best_5y'),
+	                                  sqlCol(indicator.name + '_5y', iif(indicator.is_ASC, max, min), 'worst_5y'),
+	                                  sqlCol(indicator.name + '_10y', mean, 'avg_10y'), sqlCol(indicator.name + '_10y', count, 'avg_10y_cnt'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_10y'),
+	                                  sqlCol(indicator.name + '_10y', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_10y'),
+	                                  sqlCol(indicator.name + '_10y', iif(indicator.is_ASC, min, max), 'best_10y'),
+	                                  sqlCol(indicator.name + '_10y', iif(indicator.is_ASC, max, min), 'worst_10y'),
+	                                  sqlCol(indicator.name + '_ytd', mean, 'avg_ytd'), sqlCol(indicator.name + '_ytd', count, 'avg_ytd_cnt'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', iif(indicator.is_ASC, min, max), 'best_ytd'),
+	                                  sqlCol(indicator.name + '_ytd', iif(indicator.is_ASC, max, min), 'worst_ytd')
+	                                  ),
+	                        from = cj(data_table, t_tmp),
+	                        where = <_$ranking_by IS NOT NULL>,
+	                        groupBy = (sqlCol(ranking_by), sqlCol('end_date')),
+	                        groupFlag = 1).eval(); // group by
+	}
+
+    return t_ranking, t_ranking_num;
+}
+
+
+/*
+ *   动态生成用于排序的SQL脚本
+ * 
+ *   @param indicator_name <STRING>: 指标字段名
+ *   @param indicator_id <INT>：指标ID
+ *   @param is_ASC <BOOL>: 是否排正序
+ *   @param ranking_by <STRING>: 'strategy', 'substrategy', 'factor_id', 'catavg'
  *   
+ *   TODO: portfolio, cf, manager, company, 
  *   TODO: bfi & category
  * 
  */
-def gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, ranking_by) {
-
-    // 近1月和近3月排名仅对收益有效，为了满足表结构的要求，需要建立几个”假”字段，并用NULL赋值
-    t_tmp = table(1000:0, ['indicator_id', 'indicator_1m', 'absrank_1m', 'perrank_1m',
-                                           'indicator_3m', 'absrank_3m', 'perrank_3m'],
-                          [INT, DOUBLE, INT, INT, DOUBLE, INT, INT]);
-    INSERT INTO t_tmp VALUES (indicator_id, double(NULL), int(NULL), int(NULL), double(NULL), int(NULL), int(NULL));
-
-    // 因为 parseExpr 没法将表 data_table 传入，所以用 sql()
-    t_ranking = sql(select = (sqlCol('entity_id'), sqlCol('end_date'), sqlCol(ranking_by), sqlCol('indicator_id'),
-                              sqlCol('indicator_1m'), sqlCol('absrank_1m'), sqlCol('perrank_1m'),
-                              sqlCol('indicator_3m'), sqlCol('absrank_3m'), sqlCol('perrank_3m'),
-                              // 与 MySQL 不同，这里统一把近4年和成立以来的排名去掉
-                              sqlCol(indicator_name + '_6m',,'indicator_6m'),
-                              sqlCol(indicator_name + '_6m', rank{, is_ASC}, 'absrank_6m'),
-                              sqlCol(indicator_name + '_6m', perRank{, is_ASC}, 'perrank_6m'),
-                              sqlCol(indicator_name + '_1y',,'indicator_1y'),
-                              sqlCol(indicator_name + '_1y', rank{, is_ASC}, 'absrank_1y'),
-                              sqlCol(indicator_name + '_1y', perRank{, is_ASC}, 'perrank_1y'),
-                              sqlCol(indicator_name + '_2y',,'indicator_2y'),
-                              sqlCol(indicator_name + '_2y', rank{, is_ASC}, 'absrank_2y'),
-                              sqlCol(indicator_name + '_2y', perRank{, is_ASC}, 'perrank_2y'),
-                              sqlCol(indicator_name + '_3y',,'indicator_3y'),
-                              sqlCol(indicator_name + '_3y', rank{, is_ASC}, 'absrank_3y'),
-                              sqlCol(indicator_name + '_3y', perRank{, is_ASC}, 'perrank_3y'),
-                              sqlCol(indicator_name + '_5y',,'indicator_5y'),
-                              sqlCol(indicator_name + '_5y', rank{, is_ASC}, 'absrank_5y'),
-                              sqlCol(indicator_name + '_5y', perRank{, is_ASC}, 'perrank_5y'),
-                              sqlCol(indicator_name + '_10y',,'indicator_10y'),
-                              sqlCol(indicator_name + '_10y', rank{, is_ASC}, 'absrank_10y'),
-                              sqlCol(indicator_name + '_10y', perRank{, is_ASC}, 'perrank_10y'),
-                              sqlCol(indicator_name + '_ytd',,'indicator_ytd'),
-                              sqlCol(indicator_name + '_ytd', rank{, is_ASC}, 'absrank_ytd'),
-                              sqlCol(indicator_name + '_ytd', perRank{, is_ASC}, 'perrank_ytd')
-                              ), 
-                    from = cj(data_table, t_tmp),
-                    where = <_$ranking_by IS NOT NULL>,
-                    groupBy = (sqlCol(ranking_by), sqlCol('end_date')),
-                    groupFlag = 0 ).eval(); // context by
-
-
-    // 近1月和近3月排名仅对收益有效，为了满足表结构的要求，需要建立几个”假”字段，并用NULL赋值
-    t_tmp = table(1000:0, ['indicator_id', 'avg_1m', 'avg_1m_cnt', 'perrank_percent_5_1m', 'perrank_percent_10_1m', 'perrank_percent_25_1m',
-                           'perrank_percent_50_1m', 'perrank_percent_75_1m', 'perrank_percent_90_1m', 'perrank_percent_95_1m', 'best_1m', 'worst_1m',
-                           'avg_3m', 'avg_3m_cnt', 'perrank_percent_5_3m', 'perrank_percent_10_3m', 'perrank_percent_25_3m',
-                           'perrank_percent_50_3m', 'perrank_percent_75_3m', 'perrank_percent_90_3m', 'perrank_percent_95_3m', 'best_3m', 'worst_3m'],
-                          [INT, DOUBLE, INT, DOUBLE, DOUBLE, DOUBLE,
-                           DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE,
-                           DOUBLE, INT, DOUBLE, DOUBLE, DOUBLE,
-                           DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE, DOUBLE]);
-
-    INSERT INTO t_tmp VALUES (indicator_id, double(NULL), int(NULL), double(NULL), double(NULL), double(NULL),
-                              double(NULL), double(NULL), double(NULL), double(NULL), double(NULL), double(NULL),
-                              double(NULL), int(NULL), double(NULL), double(NULL), double(NULL),
-                              double(NULL), double(NULL), double(NULL), double(NULL), double(NULL), double(NULL)); 
-
-    t_ranking_num = sql(select = (sqlCol('end_date'), sqlCol(ranking_by), sqlCol('raise_type', mean, 'raise_type'), sqlCol('indicator_id', mean,'indicator_id'),
-                                  sqlCol('avg_1m', mean, 'avg_1m'), sqlCol('avg_1m_cnt', mean, 'avg_1m_cnt'),
-                                  sqlCol('perrank_percent_5_1m', mean, 'perrank_percent_5_1m'),
-                                  sqlCol('perrank_percent_10_1m', mean, 'perrank_percent_10_1m'),
-                                  sqlCol('perrank_percent_25_1m', mean, 'perrank_percent_25_1m'),
-                                  sqlCol('perrank_percent_50_1m', mean, 'perrank_percent_50_1m'),
-                                  sqlCol('perrank_percent_75_1m', mean, 'perrank_percent_75_1m'),
-                                  sqlCol('perrank_percent_90_1m', mean, 'perrank_percent_90_1m'),
-                                  sqlCol('perrank_percent_95_1m', mean, 'perrank_percent_95_1m'), 
-                                  sqlCol('best_1m', mean, 'best_1m'), sqlCol('worst_1m', mean, 'worst_1m'),
-                                  sqlCol('avg_3m', mean, 'avg_3m'), sqlCol('avg_3m_cnt', mean, 'avg_3m_cnt'),
-                                  sqlCol('perrank_percent_5_3m', mean, 'perrank_percent_5_3m'),
-                                  sqlCol('perrank_percent_10_3m', mean, 'perrank_percent_10_3m'),
-                                  sqlCol('perrank_percent_25_3m', mean, 'perrank_percent_25_3m'),
-                                  sqlCol('perrank_percent_50_3m', mean, 'perrank_percent_50_3m'),
-                                  sqlCol('perrank_percent_75_3m', mean, 'perrank_percent_75_3m'),
-                                  sqlCol('perrank_percent_90_3m', mean, 'perrank_percent_90_3m'),
-                                  sqlCol('perrank_percent_95_3m', mean, 'perrank_percent_95_3m'), 
-                                  sqlCol('best_3m', mean, 'best_3m'), sqlCol('worst_3m', mean, 'worst_3m'),
-                                  // 与 MySQL 不同，这里统一把近4年和成立以来的排名去掉
-                                  sqlCol(indicator_name + '_6m', mean, 'avg_6m'), sqlCol(indicator_name + '_6m', count, 'avg_6m_cnt'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_6m'),
-                                  sqlCol(indicator_name + '_6m', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_6m'),
-                                  sqlCol(indicator_name + '_6m', iif(is_ASC, min, max), 'best_6m'),
-                                  sqlCol(indicator_name + '_6m', iif(is_ASC, max, min), 'worst_6m'),
-                                  sqlCol(indicator_name + '_1y', mean, 'avg_1y'), sqlCol(indicator_name + '_1y', count, 'avg_1y_cnt'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_1y'),
-                                  sqlCol(indicator_name + '_1y', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_1y'),
-                                  sqlCol(indicator_name + '_1y', iif(is_ASC, min, max), 'best_1y'),
-                                  sqlCol(indicator_name + '_1y', iif(is_ASC, max, min), 'worst_1y'),
-                                  sqlCol(indicator_name + '_2y', mean, 'avg_2y'), sqlCol(indicator_name + '_2y', count, 'avg_2y_cnt'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_2y'),
-                                  sqlCol(indicator_name + '_2y', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_2y'),
-                                  sqlCol(indicator_name + '_2y', iif(is_ASC, min, max), 'best_2y'),
-                                  sqlCol(indicator_name + '_2y', iif(is_ASC, max, min), 'worst_2y'),
-                                  sqlCol(indicator_name + '_3y', mean, 'avg_3y'), sqlCol(indicator_name + '_3y', count, 'avg_3y_cnt'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_3y'),
-                                  sqlCol(indicator_name + '_3y', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_3y'),
-                                  sqlCol(indicator_name + '_3y', iif(is_ASC, min, max), 'best_3y'),
-                                  sqlCol(indicator_name + '_3y', iif(is_ASC, max, min), 'worst_3y'),
-                                  sqlCol(indicator_name + '_5y', mean, 'avg_5y'), sqlCol(indicator_name + '_5y', count, 'avg_5y_cnt'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_5y'),
-                                  sqlCol(indicator_name + '_5y', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_5y'),
-                                  sqlCol(indicator_name + '_5y', iif(is_ASC, min, max), 'best_5y'),
-                                  sqlCol(indicator_name + '_5y', iif(is_ASC, max, min), 'worst_5y'),
-                                  sqlCol(indicator_name + '_10y', mean, 'avg_10y'), sqlCol(indicator_name + '_10y', count, 'avg_10y_cnt'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_10y'),
-                                  sqlCol(indicator_name + '_10y', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_10y'),
-                                  sqlCol(indicator_name + '_10y', iif(is_ASC, min, max), 'best_10y'),
-                                  sqlCol(indicator_name + '_10y', iif(is_ASC, max, min), 'worst_10y'),
-                                  sqlCol(indicator_name + '_ytd', mean, 'avg_ytd'), sqlCol(indicator_name + '_ytd', count, 'avg_ytd_cnt'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 5, 95)}, 'perrank_percent_5_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 10, 90)}, 'perrank_percent_10_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 25, 75)}, 'perrank_percent_25_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 50, 50)}, 'perrank_percent_50_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 75, 25)}, 'perrank_percent_75_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 90, 10)}, 'perrank_percent_90_ytd'),
-                                  sqlCol(indicator_name + '_ytd', percentile{, iif(is_ASC, 95, 5)}, 'perrank_percent_95_ytd'),
-                                  sqlCol(indicator_name + '_ytd', iif(is_ASC, min, max), 'best_ytd'),
-                                  sqlCol(indicator_name + '_ytd', iif(is_ASC, max, min), 'worst_ytd')
-                                  ),
-                        from = cj(data_table, t_tmp),
-                        where = <_$ranking_by IS NOT NULL>,
-                        groupBy = (sqlCol(ranking_by), sqlCol('end_date')),
-                        groupFlag = 1).eval(); // group by
+def gen_ranking_sql(data_table, indicator_table, ranking_by) {
+
+	for(indicator in indicator_table) {
+
+        // 与 MySQL 不同，这里统一把近4年和成立以来的排名去掉
+		if(indicator.id == 1)
+			v_trailing = ['1m', '3m', '6m', '1y', '2y', '3y', '5y', '10y', 'ytd'];
+		else {
+			v_trailing = ['6m', '1y', '2y', '3y', '5y', '10y', 'ytd'];
+			v_missing_trailing = ['1m', '3m'];
+		}
+
+        // 绝对排名和百分位排名
+	    t_ranking = sql(select = (sqlCol(['entity_id', 'end_date', 'category_id']), <indicator.id as indicator_id>,
+	                              sqlCol(indicator.name + '_' + v_trailing,, 'indicator_' + v_trailing),
+	                              sqlCol(indicator.name + '_' + v_trailing, rank{, indicator.is_ASC}, 'absrank_' + v_trailing),
+	                              sqlCol(indicator.name + '_' + v_trailing, perRank{, indicator.is_ASC}, 'perrank_' + v_trailing)
+                                 ), 
+	                    from = data_table,
+	                    where = < category_id IS NOT NULL>,
+	                    groupBy = sqlCol(['category_id', 'end_date']),
+	                    groupFlag = 0 ).eval(); // context by
+
+	    // 为了满足表结构的要求, 非收益的指标要补上1m和3m的字段，虽然都是NULL
+	    if(indicator.id != 1) {
+
+	    	t_tmp = table(1000:0,
+	    	              ['indicator_' + v_missing_trailing, 'absrank_' + v_missing_trailing, 'perrank_' + v_missing_trailing].flatten(),
+	    	              [take(DOUBLE, v_missing_trailing.size()), take(INT, v_missing_trailing.size()), take(INT, v_missing_trailing.size())].flatten()
+	    	             );
+	    	
+	    	t_ranking = SELECT * FROM cj(t_ranking, t_tmp);
+	    }
+
+	    // 平均值、集合数量、各分位的阈值
+        t_ranking_num = sql(select =(sqlCol(['end_date', 'category_id']),
+                                     sqlCol('raise_type', mean, 'raise_type'), <indicator.id as indicator_id>,
+                                     sqlCol(indicator.name + '_' + v_trailing, mean, 'avg_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, count, 'avg_' + v_trailing + '_cnt'),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 5, 95)}, 'perrank_percent_5_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 10, 90)}, 'perrank_percent_10_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 25, 75)}, 'perrank_percent_25_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 50, 50)}, 'perrank_percent_50_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 75, 25)}, 'perrank_percent_75_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 90, 10)}, 'perrank_percent_90_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, percentile{, iif(indicator.is_ASC, 95, 5)}, 'perrank_percent_95_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, iif(indicator.is_ASC, min, max), 'best_' + v_trailing),
+                                     sqlCol(indicator.name + '_' + v_trailing, iif(indicator.is_ASC, max, min), 'worst_' + v_trailing)
+                                    ),
+        	                 from = data_table,
+        	                 where = < category_id IS NOT NULL>,
+        	                 groupBy = sqlCol(['category_id', 'end_date']),
+        	                 groupFlag = 1).eval(); // group by
+
+	    // 为了满足表结构的要求, 非收益的指标要补上1m和3m的字段，虽然都是NULL
+	    if(indicator.id != 1) {
+
+	    	t_tmp = table(1000:0,
+	    	              ['avg_' + v_missing_trailing, 'avg_' + v_missing_trailing + '_cnt', 'perrank_percent_5_' + v_missing_trailing,
+	    	               'perrank_percent_10_' + v_missing_trailing, 'perrank_percent_25_' + v_missing_trailing,
+	    	               'perrank_percent_50_' + v_missing_trailing, 'perrank_percent_75_' + v_missing_trailing,
+	    	               'perrank_percent_90_' + v_missing_trailing, 'perrank_percent_95_' + v_missing_trailing,
+	    	               'best_' + v_missing_trailing, 'worst_' + v_missing_trailing ].flatten(),
+	    	              [take(DOUBLE, v_missing_trailing.size()), take(INT, v_missing_trailing.size()), take(DOUBLE, v_missing_trailing.size()),
+	    	               take(DOUBLE, v_missing_trailing.size()), take(DOUBLE, v_missing_trailing.size()),
+	    	               take(DOUBLE, v_missing_trailing.size()), take(DOUBLE, v_missing_trailing.size()),
+	    	               take(DOUBLE, v_missing_trailing.size()), take(DOUBLE, v_missing_trailing.size()),
+	    	               take(DOUBLE, v_missing_trailing.size()),take(DOUBLE, v_missing_trailing.size())].flatten()
+	    	             );
+	    	
+	    	t_ranking = SELECT * FROM cj(t_ranking, t_tmp);
+	    }
+	}
 
     return t_ranking, t_ranking_num;
 }
@@ -343,18 +472,18 @@ def gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, ranking_by
  * 
  *  NOTE: 没有用 parseExpr 来生成动态脚本的原因是数据表无法传入
  */
-def run_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, mutable v_tables) {
+def run_ranking_sql(data_table, indicator_table, mutable v_tables) {
 
-    tb_strategy_ranking = gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, 'strategy')[0];
+    tb_strategy_ranking = gen_ranking_sql(data_table, indicator_table, 'strategy')[0];
     v_tables[0].tableInsert(tb_strategy_ranking);
 
-    tb_strategy_ranking_num = gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, 'strategy')[1];
+    tb_strategy_ranking_num = gen_ranking_sql(data_table, indicator_table, 'strategy')[1];
     v_tables[1].tableInsert(tb_strategy_ranking_num);
 
-    tb_substrategy_ranking = gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, 'substrategy')[0];
+    tb_substrategy_ranking = gen_ranking_sql(data_table, indicator_table, 'substrategy')[0];
     v_tables[2].tableInsert(tb_substrategy_ranking);
 
-    tb_substrategy_ranking_num = gen_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, 'substrategy')[1];
+    tb_substrategy_ranking_num = gen_ranking_sql(data_table, indicator_table, 'substrategy')[1];
     v_tables[3].tableInsert(tb_substrategy_ranking_num);
 
 }
@@ -366,6 +495,9 @@ def run_ranking_sql(data_table, indicator_name, indicator_id, is_ASC, mutable v_
  */
 def cal_risk_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
+
 	table_desc = get_risk_stats_table_description(entity_type);
 
 	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
@@ -385,35 +517,10 @@ def cal_risk_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
     v_ranking_tables = [t_s, t_s_num, t_ss, t_ss_num];
 
-    // 最大回撤
-    run_ranking_sql(t, 'maxdrawdown', 2, true, v_ranking_tables);
-
-    // 峰度
-    run_ranking_sql(t, 'kurtosis', 6, true, v_ranking_tables);
+    // 50, 52 dolphin 未计算
+    indicator_table = SELECT * FROM get_indicator_info() WHERE id in [2, 6, 9, 10, 11, 12, 21, 59];
 
-    // 偏度
-    run_ranking_sql(t, 'skewness', 9, false, v_ranking_tables);
-
-    // 标准差
-    run_ranking_sql(t, 'stddev', 10, true, v_ranking_tables);
-
-    // Alpha
-    run_ranking_sql(t, 'alpha', 11, false, v_ranking_tables);
-
-    // Beta
-    run_ranking_sql(t, 'beta', 12, false, v_ranking_tables);
-
-    // 下行标准差
-    run_ranking_sql(t, 'downsidedev', 21, true, v_ranking_tables);
-
-    // 月最大回撤  dolphin 未计算
-    // run_ranking_sql(t, 'maxdrawdown_months', 50, true, v_ranking_tables);
-
-    // 最大回撤修复月份数   dolphin 未计算
-    //run_ranking_sql(t, 'maxdrawdown_recoverymonths', 52, true, v_ranking_tables);
-    
-    // 胜率
-    run_ranking_sql(t, 'winrate', 59, false, v_ranking_tables);
+    run_ranking_sql(t, indicator_table, v_ranking_tables);
 
     return v_ranking_tables;
 }
@@ -426,6 +533,9 @@ def cal_risk_ranking(entity_type, entity_info, end_date, isFromMySQL) {
  */
 def cal_risk_adj_return_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
+
 	table_desc = get_riskadjret_stats_table_description(entity_type);
 
 	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
@@ -445,42 +555,64 @@ def cal_risk_adj_return_ranking(entity_type, entity_info, end_date, isFromMySQL)
 
     v_ranking_tables = [t_s, t_s_num, t_ss, t_ss_num];
 
-    // Kappa ratio
-    run_ranking_sql(t, 'kapparatio', 14, false, v_ranking_tables);
+    // 19 (MAR Sortino ratio) dolphin 未计算
+    indicator_table = SELECT * FROM get_indicator_info() WHERE id in [14, 15, 16, 17, 18, 40, 58];
 
-    // Treynor ratio
-    run_ranking_sql(t, 'treynorratio', 15, false, v_ranking_tables);
+    run_ranking_sql(t, indicator_table, v_ranking_tables);
 
-    // Jensen
-    run_ranking_sql(t, 'jensen', 16, false, v_ranking_tables);
+    return v_ranking_tables;
+}
 
-    // Omega ratio
-    run_ranking_sql(t, 'omegaratio', 17, false, v_ranking_tables);
 
-    // Sharpe ratio
-    run_ranking_sql(t, 'sharperatio', 18, false, v_ranking_tables);
+/*
+ *   计算杂项指标排名
+ * 
+ * 
+ */
+def cal_other_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL) {
+
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
 
-    // MAR Sortino ratio   dolphin 未计算
-    //run_ranking_sql(t, 'sortinoratio_MAR', 19, false, v_ranking_tables);
+	table_desc = get_indicator_table_description(entity_type);
 
-    // Calmar ratio
-    run_ranking_sql(t, 'calmarratio', 40, false, v_ranking_tables);
+	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data.rename!(sec_id_col, 'entity_id');
 
-    // Sortino ratio
-    run_ranking_sql(t, 'sortinoratio', 58, false, v_ranking_tables);
+	t = SELECT *
+        FROM entity_info en 
+        INNER JOIN tb_data d ON en.entity_id = d.entity_id
+	    WHERE en.strategy IS NOT NULL;
+
+    // 按照 MySQL 字段建表
+    t_s = create_entity_indicator_ranking(false);
+    t_s_num = create_entity_indicator_ranking_num(false);
+    t_ss = create_entity_indicator_substrategy_ranking(false);
+    t_ss_num = create_entity_indicator_substrategy_ranking_num(false);
+
+    v_ranking_tables = [t_s, t_s_num, t_ss, t_ss_num];
+
+    // 37 (per_con), 43, 44, 45, 46, 47 (smdd模型) dolphin 未计算
+    indicator_table = SELECT * FROM get_indicator_info() WHERE id in [38, 41, 42, 48, 49];
+
+    run_ranking_sql(t, indicator_table, v_ranking_tables);
 
     return v_ranking_tables;
 }
 
 
 /*
- *   计算杂项指标排名
+ *   计算上下行指标排名
  * 
  * 
  */
-def cal_other_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL) {
+def cal_capture_style_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
-	table_desc = get_indicator_table_description(entity_type);
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
+
+	table_desc = get_capture_style_table_description(entity_type);
 
 	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
 	sec_id_col = table_desc.sec_id_col[0];
@@ -499,43 +631,73 @@ def cal_other_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL)
 
     v_ranking_tables = [t_s, t_s_num, t_ss, t_ss_num];
 
-    // 风格一致性  dolphin 未计算
-    //run_ranking_sql(t, 'per_con', 37, false, v_ranking_tables);
+    indicator_table = SELECT * FROM get_indicator_info() WHERE id in [33, 34, 35, 36];
 
-    // Information ratio
-    run_ranking_sql(t, 'info_ratio', 38, false, v_ranking_tables);
+    run_ranking_sql(t, indicator_table, v_ranking_tables);
 
-    // Value at Risk
-    run_ranking_sql(t, 'var', 41, true, v_ranking_tables);
+    return v_ranking_tables;
+}
 
-    // Conditional Value at Risk
-    run_ranking_sql(t, 'cvar', 42, true, v_ranking_tables);
 
-    // SMDD 模型下的 VaR  dolphin 未计算
-    //run_ranking_sql(t, 'smddvar', 43, true, v_ranking_tables);
+/*
+ *   计算BFI指标排名
+ * 
+ *   TODO: return
+ */
+def cal_bfi_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL) {
 
-    // SMDD 模型下的 CVaR  dolphin 未计算
-    //run_ranking_sql(t, 'smddcvar', 44, true, v_ranking_tables);
+    // 当前只对基金做排名， 其它类型参考基金排名做相对排名
+    if(!(entity_type in ['MF', 'HF'])) return null;
 
-    // SMDD 模型下的 LPM1  dolphin 未计算
-    //run_ranking_sql(t, 'smdd_lpm1', 45, true, v_ranking_tables);
-    
-    // SMDD 模型下的 LPM2  dolphin 未计算
-    //run_ranking_sql(t, 'smdd_lpm2', 46, true, v_ranking_tables);
-    
-    // SMDD 模型下的下行风险  dolphin 未计算
-    //run_ranking_sql(t, 'smdd_downside_dev', 47, true, v_ranking_tables);
+	table_desc = get_bfi_indicator_table_description(entity_type);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data_bfi_indicator = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	tb_data_bfi_indicator.rename!(sec_id_col, 'entity_id');
+
+	table_desc = get_risk_stats_table_description(entity_type);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data_risk_stats = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	tb_data_risk_stats.rename!(sec_id_col, 'entity_id');
+
+	table_desc = get_riskadjret_stats_table_description(entity_type);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data_riskadjret_stats = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	tb_data_riskadjret_stats.rename!(sec_id_col, 'entity_id');
+	
+	table_desc = get_indicator_table_description(entity_type);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data_indicator_stats = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	tb_data_indicator_stats.rename!(sec_id_col, 'entity_id');
+
+	t = SELECT *
+        FROM entity_info en 
+        INNER JOIN tb_data_bfi_indicator d2 ON en.entity_id = d2.entity_id
+        INNER JOIN tb_data_risk_stats d3 ON en.entity_id = d3.entity_id
+        INNER JOIN tb_data_riskadjret_stats d4 ON en.entity_id = d4.entity_id
+        INNER JOIN tb_data_indicator_stats d5 ON en.entity_id = d5.entity_id
+	    WHERE en.strategy IS NOT NULL;
     
-    // 跟踪误差
-    run_ranking_sql(t, 'tracking_error', 48, true, v_ranking_tables);
+    // 按照 MySQL 字段建表
+    t_s = create_entity_indicator_ranking(false);
+    t_s_num = create_entity_indicator_ranking_num(false);
+    t_ss = create_entity_indicator_substrategy_ranking(false);
+    t_ss_num = create_entity_indicator_substrategy_ranking_num(false);
+
+    v_ranking_tables = [t_s, t_s_num, t_ss, t_ss_num];
 
-    // M2
-    run_ranking_sql(t, 'm2', 49, false, v_ranking_tables);
+    // 取消 39, 53, 54, 55, 57, 57
+    v_indicator_id = [11, 12, 16, 33, 34, 35, 36, 38, 48, 59,
+                      2, 6, 9, 10, 21, 50, 52,
+                      14, 15, 16, 17, 18, 19, 58, 21, 40,
+                      37, 41, 42, 43, 44, 45, 46, 47, 49
+                     ];
+    indicator_table = SELECT * FROM get_indicator_info() WHERE id in v_indicator_id;
+
+    run_ranking_sql(t, indicator_table, v_ranking_tables);
 
     return v_ranking_tables;
 }
 
-
 /*
  *   将源指标表横表变竖表，以方便排名计算
  * 
@@ -543,7 +705,11 @@ def cal_other_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL)
  */
 def transform_data_for_ranking(entity_type, data_table, ranking_by, indicator_info) {
 
-    v_trailing = ['6m', '1y', '2y', '3y', '5y', '10y', 'ytd'];
+    // 只有收益需要1m, 3m
+    if(indicator_info.size() == 1 && indicator_info[0].id == 1 )
+		v_trailing = ['1m', '3m', '6m', '1y', '2y', '3y', '5y', '10y', 'ytd'];
+	else
+	    v_trailing = ['6m', '1y', '2y', '3y', '5y', '10y', 'ytd'];
 
     // 只有 portfolio_id 是整型，其它的都是字符串
     is_id_integer = false;
@@ -558,35 +724,38 @@ def transform_data_for_ranking(entity_type, data_table, ranking_by, indicator_in
 
     for(indicator in indicator_info) {
 
-        t = sql(select = (sqlCol('entity_id'), sqlCol('end_date'), sqlCol('category_id'),
+        t = sql(select = (sqlCol(['entity_id', 'end_date', 'category_id']), <indicator.id as indicator_id>,
                           sqlCol(indicator.name + '_' + v_trailing,, 'indicator_' + v_trailing)
                          ),
                 from = data_table
         	   ).eval();
 
-        t.join!(table(take(indicator.id, t.size()) AS indicator_id,
-                      take(double(NULL), t.size()) AS indicator_1m,
-                      take(int(NULL), t.size()) AS absrank_1m,
-                      take(int(NULL), t.size()) AS perrank_1m,
-                      take(double(NULL), t.size()) AS indicator_3m,
-                      take(int(NULL), t.size()) AS absrank_3m,
-                      take(int(NULL), t.size()) AS perrank_3m,
-                      take(int(NULL), t.size()) AS absrank_6m,
-                      take(int(NULL), t.size()) AS perrank_6m,
-                      take(int(NULL), t.size()) AS absrank_1y,
-                      take(int(NULL), t.size()) AS perrank_1y,
-                      take(int(NULL), t.size()) AS absrank_2y,
-                      take(int(NULL), t.size()) AS perrank_2y,
-                      take(int(NULL), t.size()) AS absrank_3y,
-                      take(int(NULL), t.size()) AS perrank_3y,
-                      take(int(NULL), t.size()) AS absrank_5y,
-                      take(int(NULL), t.size()) AS perrank_5y,
-                      take(int(NULL), t.size()) AS absrank_10y,
-                      take(int(NULL), t.size()) AS perrank_10y,
-                      take(int(NULL), t.size()) AS absrank_ytd,
-                      take(int(NULL), t.size()) AS perrank_ytd)
-                );
+        if(indicator_info.size() > 1 || indicator_info[0].id <> 1 )
+        {
+	        t.join!(table(take(double(NULL), t.size()) AS indicator_1m,
+                          take(double(NULL), t.size()) AS indicator_3m) );
+        }
 
+        t.join!(table( take(int(NULL), t.size()) AS absrank_1m,
+                       take(int(NULL), t.size()) AS perrank_1m,
+                       take(int(NULL), t.size()) AS absrank_3m,
+                       take(int(NULL), t.size()) AS perrank_3m,
+                       take(int(NULL), t.size()) AS absrank_6m,
+                       take(int(NULL), t.size()) AS perrank_6m,
+                       take(int(NULL), t.size()) AS absrank_1y,
+                       take(int(NULL), t.size()) AS perrank_1y,
+                       take(int(NULL), t.size()) AS absrank_2y,
+                       take(int(NULL), t.size()) AS perrank_2y,
+                       take(int(NULL), t.size()) AS absrank_3y,
+                       take(int(NULL), t.size()) AS perrank_3y,
+                       take(int(NULL), t.size()) AS absrank_5y,
+                       take(int(NULL), t.size()) AS perrank_5y,
+                       take(int(NULL), t.size()) AS absrank_10y,
+                       take(int(NULL), t.size()) AS perrank_10y,
+                       take(int(NULL), t.size()) AS absrank_ytd,
+                       take(int(NULL), t.size()) AS perrank_ytd )
+    	       );
+       
         INSERT INTO tb_ranking
             SELECT * FROM (sql(select = sqlCol(tb_ranking.colNames()),
                                from = t).eval());
@@ -599,6 +768,39 @@ def transform_data_for_ranking(entity_type, data_table, ranking_by, indicator_in
 /*
  *   将源风险指标表横表变竖表，以方便排名计算
  * 
+ *   
+ */
+def transform_return_for_ranking (entity_type, entity_info, end_date, ranking_by, isFromMySQL=true) {
+
+	table_desc = get_performance_table_description(entity_type);
+
+	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data.rename!(sec_id_col, 'entity_id');
+
+	data_table = SELECT * FROM entity_info en 
+                 INNER JOIN tb_data d ON en.entity_id = d.entity_id
+	             WHERE en.strategy IS NOT NULL
+
+    data_table.dropColumns!('id');
+    data_table.rename!(ranking_by, 'category_id');
+
+
+    // 目前SQL排名的指标还包含39：年化收益，与收益没有什么差别所以被取消
+    v_indicator_name = ['ret'];
+    v_indicator_id = [1];
+    v_is_ASC = [false];
+    t_indicator = table(v_indicator_name AS name, v_indicator_id AS id, v_is_ASC AS is_ASC);
+
+    tb_ranking = transform_data_for_ranking(entity_type, data_table, ranking_by, t_indicator).rename!('category_id', ranking_by);
+
+    return tb_ranking;
+}
+
+
+/*
+ *   将源风险指标表横表变竖表，以方便排名计算
+ * 
  *   TODO: 一直缺 portfolio bfi indicator 计算！mysql 里的 pf_fund_bfi_bm_indicator_ranking 是错的... 
  */
 def transform_risk_stats_for_ranking (entity_type, entity_info, end_date, ranking_by, isFromMySQL=true) {
@@ -692,6 +894,38 @@ def transform_other_indicator_for_ranking (entity_type, entity_info, end_date, r
     return tb_ranking;
 }
 
+
+/*
+ *   将源杂项风险指标表横表变竖表，以方便排名计算
+ * 
+ * 
+ */
+def transform_capture_style_for_ranking (entity_type, entity_info, end_date, ranking_by, isFromMySQL=true) {
+
+	table_desc = get_capture_style_table_description(entity_type);
+
+	tb_data = get_monthly_indicator_data(table_desc.table_name[0], end_date, isFromMySQL);
+	sec_id_col = table_desc.sec_id_col[0];
+	tb_data.rename!(sec_id_col, 'entity_id');
+
+	data_table = SELECT * FROM entity_info en 
+                 INNER JOIN tb_data d ON en.entity_id = d.entity_id
+	             WHERE en.strategy IS NOT NULL
+
+    data_table.dropColumns!('id');
+    data_table.rename!(ranking_by, 'category_id');
+
+    // 目前SQL排名的指标
+    v_indicator_name = ['upsidecapture_ret', 'downsidecapture_ret', 'upsidecapture_ratio', 'downsidecapture_ratio'];
+    v_indicator_id = [33, 34, 35, 36];
+    v_is_ASC = [false, false, false, true];
+    t_indicator = table(v_indicator_name AS name, v_indicator_id AS id, v_is_ASC AS is_ASC);
+
+    tb_ranking = transform_data_for_ranking(entity_type, data_table, ranking_by, t_indicator).rename!('category_id', ranking_by);
+
+    return tb_ranking;
+}
+
 /*
  * 
  *   参考某指定类排名，计算相对排名
@@ -778,15 +1012,19 @@ def save_ranking_tables(entity_type, ranking_tables) {
 
     save_and_sync(ranking_tables[1], source_table + '_num', target_table + '_num');
     
+    source_table = source_table.strReplace('_ranking', '_substrategy_ranking');
+    target_table = target_table.strReplace('_ranking', '_substrategy_ranking');
+
     ranking_tables[2].rename!('entity_id', entity_id_col);
-    save_and_sync(ranking_tables[2], source_table.replace!('_ranking', '_substrategy_ranking'), target_table.replace!('_ranking', '_substrategy_ranking'));
+    save_and_sync(ranking_tables[2], source_table, target_table);
 
     save_and_sync(ranking_tables[3], source_table + '_num', target_table + '_num');
 	
 }
 
 /*
- *  参考排名数据入库
+ *  参考排名数据入库AND a.indicator_id NOT IN (50, 52, 59, 46)
+
  * 
  *  @param ranking_tables <TABLE>: 
  */

modules/sqlUtilities.dos

@@ -186,6 +186,50 @@ def get_riskadjret_stats_table_description(entity_type) {
 
 }
 
+
+/*
+ *  根据不同类型的主体返回其杂项指标的表名、字段名
+ * 
+ *  Example: get_capture_style_table_description('MF');
+ */
+def get_capture_style_table_description(entity_type) {
+
+    tmp_universe = table(100:0, 
+                         ['type', 'table_name', 'sec_id_col'],
+                         [STRING, STRING, STRING]);
+
+    // 分别对应：私募，公募，私有基金，市场指数，图译指数，私有指数，图译因子，组合
+    INSERT INTO tmp_universe VALUES ( 
+        ['HF', 'MF', 'CF', 'MI', 'FI', 'CI', 'FA', 'PF'],
+        ['mfdb.fund_style_stats', 'mfdb.fund_style_stats', 'pfdb.pf_cus_fund_style_stats', 'mfdb.fund_style_stats', 'mfdb.fund_style_stats', 'pfdb.cm_udf_index_style_stats', 'pfdb.cm_factor_style_stats', 'pfdb.pf_portfolio_style_stats'],
+        ['fund_id', 'fund_id', 'fund_id', 'fund_id', 'fund_id', 'index_id', 'factor_id', 'portfolio_id'] );
+
+    return (SELECT * FROM tmp_universe u WHERE u.type = entity_type);
+
+}
+
+
+/*
+ *  根据不同类型的主体返回其BFI指标的表名、字段名
+ * 
+ *  Example: get_risk_stats_table_description('HF');
+ */
+def get_bfi_indicator_table_description(entity_type) {
+
+    tmp_universe = table(100:0, 
+                         ['type', 'table_name', 'sec_id_col'],
+                         [STRING, STRING, STRING]);
+
+    // 分别对应：私募，公募，私有基金，市场指数，图译指数，私有指数，图译因子，组合
+    INSERT INTO tmp_universe VALUES ( 
+        ['HF', 'MF', 'CF', 'MI', 'FI', 'CI', 'FA', 'PF'],
+        ['mfdb.fund_ty_bfi_bm_indicator', 'mfdb.fund_ty_bfi_bm_indicator', NULL, NULL, NULL, NULL, NULL, 'pfdb.pf_portfolio_ty_bfi_bm_indicator'],
+        ['fund_id', 'fund_id', NULL, NULL, NULL, NULL, NULL, 'portfolio_id'] );
+
+    return (SELECT * FROM tmp_universe u WHERE u.type = entity_type);
+
+}
+
 /*
  *  Annulized multiple
  */

modules/task_monthlyPerformance.dos

@@ -41,40 +41,29 @@ isFromMySQL = true;
     v_ranking_tables = cal_other_indicator_ranking(entity_type, entity_info, end_date, isFromMySQL);
     save_ranking_tables(entity_type, v_ranking_tables);
 
-/*
-    v_tables = NULL;
-
-    if(entity_type IN ['MF', 'HF', 'MI', 'FI']) {
-    	
-    	v_tables = ['mfdb.fund_performance', 'mfdb.fund_risk_stats', 'mfdb.fund_riskadjret_stats', 'mfdb.fund_indicator', 'mfdb.fund_style_stats'];
-    	
-    } else if(entity_type = 'PF') {
-
-    	v_tables = ['pfdb.pf_portfolio_performance', 'pfdb.pf_portfolio_risk_stats', 'pfdb.pf_portfolio_riskadjret_stats', 'pfdb.pf_portfolio_indicator', 'pfdb.pf_portfolio_style_stats'];
-
-    } else if(entity_type = 'CF') {
-
-    	v_tables = ['pfdb.pf_cus_fund_performance', 'pfdb.pf_cus_fund_risk_stats', 'pfdb.pf_cus_fund_riskadjret_stats', 'pfdb.pf_cus_fund_indicator', 'pfdb.pf_cus_fund_style_stats'];
+    // 上下行捕获率排名
+    v_ranking_tables = cal_capture_style_ranking(entity_type, entity_info, end_date, isFromMySQL);
+    save_ranking_tables(entity_type, v_ranking_tables);
 
-    } else if(entity_type = 'FA') {
-    	
-    	v_tables = ['pfdb.cm_factor_performance', 'pfdb.cm_factor_risk_stats', 'pfdb.cm_factor_riskadjret_stats', 'pfdb.cm_factor_indicator', 'pfdb.cm_factor_style_stats'];
+}
 
-    } else if(entity_type = 'CI') {
-    	
-    	v_tables = ['pfdb.cm_udf_index_performance', 'pfdb.cm_udf_index_risk_stats', 'pfdb.cm_udf_index_riskadjret_stats', 'pfdb.cm_udf_index_indicator', 'pfdb.cm_udf_index_style_stats'];
-    }
+/*
+ *   Private Method: 计算相对排名并存入数据库
+ * 
+ * 
+ */
+def cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL=true) {
 
-    if(v_tables.isNull()) return null;
+    t_entity_ranking = entity_ranking;
     
-	for(table in v_tables) {
-		
-        tb_data = get_monthly_indicator_data(table, end_date, isFromMySQL);
-        
-	}
-*/
+    cal_relative_ranking(benchmark_ranking, t_entity_ranking, isFromMySQL);
+
+    t_entity_ranking.rename!('category_id', ranking_by);
+    save_relative_ranking_table(entity_type, t_entity_ranking, ranking_by);
+	
 }
 
+
 /*
  * 
  *   以公募基金为评级参考，计算组合、私有基金收益及指标排名
@@ -109,7 +98,7 @@ def CalRelativeRanking(entity_type, end_date, isFromMySQL=true) {
     		tb_fund_ranking = null;
 	    }
 
-    	if(tb_fund_ranking == null) return;
+    	if(tb_fund_ranking.isVoid() || tb_fund_ranking.size() == 0) return;
 
 	    tb_fund_ranking.rename!(ranking_by, 'category_id');
 
@@ -117,17 +106,18 @@ def CalRelativeRanking(entity_type, end_date, isFromMySQL=true) {
 	    v_indicator_id = [1];
     	benchmark_ranking = SELECT * FROM tb_fund_ranking WHERE indicator_id IN v_indicator_id;
 
-    
+    	entity_ranking = transform_return_for_ranking(entity_type, entity_info, end_date, ranking_by, isFromMySQL).rename!(ranking_by, 'category_id');
+
+        cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL);
+
+
 	    // 风险指标
     	v_indicator_id = [2, 6, 9, 10, 11, 12, 21, 50, 52, 59];
 	    benchmark_ranking = SELECT * FROM tb_fund_ranking WHERE indicator_id IN v_indicator_id;
     
 	    entity_ranking = transform_risk_stats_for_ranking(entity_type, entity_info, end_date, ranking_by, isFromMySQL).rename!(ranking_by, 'category_id');
 
-	    cal_relative_ranking(benchmark_ranking, entity_ranking, isFromMySQL);
-
-	    entity_ranking.rename!('category_id', ranking_by);
-	    save_relative_ranking_table(entity_type, entity_ranking, ranking_by);
+        cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL);
 
 
     	// 风险调整收益指标
@@ -136,10 +126,7 @@ def CalRelativeRanking(entity_type, end_date, isFromMySQL=true) {
     
 	    entity_ranking = transform_risk_adj_ret_stats_for_ranking(entity_type, entity_info, end_date, ranking_by, isFromMySQL).rename!(ranking_by, 'category_id');
 
-    	cal_relative_ranking(benchmark_ranking, entity_ranking, isFromMySQL);
-
-	    entity_ranking.rename!('category_id', ranking_by);
-	    save_relative_ranking_table(entity_type, entity_ranking, ranking_by);    
+        cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL);
 
 	    // 其它风险指标
     	v_indicator_id = [37, 38, 41, 42, 43, 44, 45, 46, 47, 48, 49];
@@ -147,10 +134,17 @@ def CalRelativeRanking(entity_type, end_date, isFromMySQL=true) {
 
     	entity_ranking = transform_other_indicator_for_ranking(entity_type, entity_info, end_date, ranking_by, isFromMySQL).rename!(ranking_by, 'category_id');
 
-	    cal_relative_ranking(benchmark_ranking, entity_ranking, isFromMySQL);
+        cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL);
+
+	    // 上下行捕获指标
+    	v_indicator_id = [33, 34, 35, 36];
+	    benchmark_ranking = SELECT * FROM tb_fund_ranking WHERE indicator_id IN v_indicator_id;
+
+    	entity_ranking = transform_capture_style_for_ranking(entity_type, entity_info, end_date, ranking_by, isFromMySQL).rename!(ranking_by, 'category_id');
+
+        cal_and_save_relative_ranking(entity_type, benchmark_ranking, entity_ranking, ranking_by, isFromMySQL);
+
 
-	    entity_ranking.rename!('category_id', ranking_by);
-	    save_relative_ranking_table(entity_type, entity_ranking, ranking_by);
     }
 //    SELECT * FROM entity_ranking WHERE entity_id = 143109