@Test public void testFilter19() throws Exception { System.out.println("testFilter19"); PigServer pigServer = new PigServer(pigContext); pigServer.registerQuery( "G1 = load " + GalaxyPair3 + " using PigStorage() as (objID1 : long, ra : float, dec : float, modelMag_u1 : float, modelMag_g1 : float, modelMag_r1 : float, modelMag_i1 : float, modelMag_z1 : float, petroR50_r1 : float, petrorad_u1 : float, petrorad_g1 : float, petrorad_r1 : float, petrorad_i1 : float, petrorad_z1 : float, petroRadErr_g1 : float, petroMag_g1 : float);"); pigServer.registerQuery( "AG1 = filter G1 by petrorad_u1 > 0.0 and petrorad_g1 > 0.0 and petrorad_r1 > 0.0 and petrorad_i1 > 0.0 and petrorad_z1 > 0.0 and petroRadErr_g1 > 0.0 and petroMag_g1 >= 16.0 and petroMag_g1 <= 21.0 and modelMag_u1 > -9999.0 and modelMag_g1 > -9999.0 and modelMag_r1 > -9999.0 and modelMag_i1 > -9999.0 and modelMag_z1 > -9999.0;"); pigServer.registerQuery( "G2 = load " + GalaxyPair4 + " using PigStorage() as (objID2 : long, ra : float, dec : float, modelMag_u2 : float, modelMag_g2 : float, modelMag_r2 : float, modelMag_i2 : float, modelMag_z2 : float, petroR50_r2 : float, petrorad_u2 : float, petrorad_g2 : float, petrorad_r2 : float, petrorad_i2 : float, petrorad_z2 : float, petroRadErr_g2: float, petroMag_g2 : float);"); pigServer.registerQuery( "AG2 = filter G2 by petrorad_u2 > 0.0 and petrorad_g2 > 0.0 and petrorad_r2 > 0.0 and petrorad_i2 > 0.0 and petrorad_z2 > 0.0 and petroRadErr_g2 > 0.0;"); Map<Operator, DataBag> derivedData = pigServer.getExamples2("AG1"); Map<Operator, DataBag> derivedData2 = pigServer.getExamples2("AG2"); assertTrue(derivedData != null); assertTrue(derivedData2 != null); } //
// org.apache.pig.test.udf.storefunc.PigPerformanceLoader() @Test public void testScriptL1() throws Exception { System.out.println("testScriptL1"); PigServer pigServer = new PigServer(pigContext); pigServer.setBatchOn(); // pigServer.registerJar("/home/kaituo/code/pig3/trunk/pigperf.jar"); pigServer.registerQuery( "A = load '" + page_viewsX.toString() + "/part-m-00000' using org.apache.pig.test.udf.storefunc.PigPerformanceLoader() as (user: chararray, action:int, timespent:int, query_term:chararray, ip_addr:long, timestamp:long,estimated_revenue:double, page_info:map[], page_links:bag{t:(p:map[])});"); // as (user, action, timespent, query_term, ip_addr, timestamp, estimated_revenue, page_info, // page_links);"); pigServer.registerQuery( "B = foreach A generate user, action, page_info, flatten(page_links) as page_links;"); // user, (int)action as action, (map[])page_info as page_info, // flatten((bag{tuple(map[])})page_links) as page_links;"); pigServer.registerQuery( "C = foreach B generate user, (action == 1 ? page_info#'a' : page_links#'b') as header;"); pigServer.registerQuery("D = group C by user parallel 40;"); pigServer.registerQuery("E = foreach D generate group, COUNT(C) as cnt;"); Map<Operator, DataBag> derivedData = pigServer.getExamples2("E"); assertTrue(derivedData != null); }