use of org.apache.solr.client.solrj.io.Tuple in project lucene-solr by apache.
the class StreamExpressionTest method testParallelMergeStream.
@Test
public void testParallelMergeStream() throws Exception {
new UpdateRequest().add(id, "0", "a_s", "hello0", "a_i", "0", "a_f", "0").add(id, "2", "a_s", "hello2", "a_i", "2", "a_f", "0").add(id, "3", "a_s", "hello3", "a_i", "3", "a_f", "3").add(id, "4", "a_s", "hello4", "a_i", "4", "a_f", "4").add(id, "1", "a_s", "hello1", "a_i", "1", "a_f", "1").add(id, "5", "a_s", "hello0", "a_i", "10", "a_f", "0").add(id, "6", "a_s", "hello2", "a_i", "8", "a_f", "0").add(id, "7", "a_s", "hello3", "a_i", "7", "a_f", "3").add(id, "8", "a_s", "hello4", "a_i", "11", "a_f", "4").add(id, "9", "a_s", "hello1", "a_i", "100", "a_f", "1").commit(cluster.getSolrClient(), COLLECTIONORALIAS);
String zkHost = cluster.getZkServer().getZkAddress();
StreamFactory streamFactory = new StreamFactory().withCollectionZkHost(COLLECTIONORALIAS, zkHost).withFunctionName("search", CloudSolrStream.class).withFunctionName("unique", UniqueStream.class).withFunctionName("top", RankStream.class).withFunctionName("group", ReducerStream.class).withFunctionName("merge", MergeStream.class).withFunctionName("parallel", ParallelStream.class);
StreamContext streamContext = new StreamContext();
SolrClientCache solrClientCache = new SolrClientCache();
streamContext.setSolrClientCache(solrClientCache);
try {
//Test ascending
ParallelStream pstream = (ParallelStream) streamFactory.constructStream("parallel(" + COLLECTIONORALIAS + ", merge(search(" + COLLECTIONORALIAS + ", q=\"id:(4 1 8 7 9)\", fl=\"id,a_s,a_i\", sort=\"a_i asc\", partitionKeys=\"a_i\"), search(" + COLLECTIONORALIAS + ", q=\"id:(0 2 3 6)\", fl=\"id,a_s,a_i\", sort=\"a_i asc\", partitionKeys=\"a_i\"), on=\"a_i asc\"), workers=\"2\", zkHost=\"" + zkHost + "\", sort=\"a_i asc\")");
pstream.setStreamContext(streamContext);
List<Tuple> tuples = getTuples(pstream);
assert (tuples.size() == 9);
assertOrder(tuples, 0, 1, 2, 3, 4, 7, 6, 8, 9);
//Test descending
pstream = (ParallelStream) streamFactory.constructStream("parallel(" + COLLECTIONORALIAS + ", merge(search(" + COLLECTIONORALIAS + ", q=\"id:(4 1 8 9)\", fl=\"id,a_s,a_i\", sort=\"a_i desc\", partitionKeys=\"a_i\"), search(" + COLLECTIONORALIAS + ", q=\"id:(0 2 3 6)\", fl=\"id,a_s,a_i\", sort=\"a_i desc\", partitionKeys=\"a_i\"), on=\"a_i desc\"), workers=\"2\", zkHost=\"" + zkHost + "\", sort=\"a_i desc\")");
pstream.setStreamContext(streamContext);
tuples = getTuples(pstream);
assert (tuples.size() == 8);
assertOrder(tuples, 9, 8, 6, 4, 3, 2, 1, 0);
} finally {
solrClientCache.close();
}
}
use of org.apache.solr.client.solrj.io.Tuple in project lucene-solr by apache.
the class StreamExpressionTest method testUniqueStream.
@Test
public void testUniqueStream() throws Exception {
new UpdateRequest().add(id, "0", "a_s", "hello0", "a_i", "0", "a_f", "0").add(id, "2", "a_s", "hello2", "a_i", "2", "a_f", "0").add(id, "3", "a_s", "hello3", "a_i", "3", "a_f", "3").add(id, "4", "a_s", "hello4", "a_i", "4", "a_f", "4").add(id, "1", "a_s", "hello1", "a_i", "1", "a_f", "1").commit(cluster.getSolrClient(), COLLECTIONORALIAS);
StreamExpression expression;
TupleStream stream;
List<Tuple> tuples;
StreamContext streamContext = new StreamContext();
SolrClientCache solrClientCache = new SolrClientCache();
streamContext.setSolrClientCache(solrClientCache);
StreamFactory factory = new StreamFactory().withCollectionZkHost(COLLECTIONORALIAS, cluster.getZkServer().getZkAddress()).withFunctionName("search", CloudSolrStream.class).withFunctionName("unique", UniqueStream.class);
try {
// Basic test
expression = StreamExpressionParser.parse("unique(search(" + COLLECTIONORALIAS + ", q=*:*, fl=\"id,a_s,a_i,a_f\", sort=\"a_f asc, a_i asc\"), over=\"a_f\")");
stream = new UniqueStream(expression, factory);
stream.setStreamContext(streamContext);
tuples = getTuples(stream);
assert (tuples.size() == 4);
assertOrder(tuples, 0, 1, 3, 4);
// Basic test desc
expression = StreamExpressionParser.parse("unique(search(" + COLLECTIONORALIAS + ", q=*:*, fl=\"id,a_s,a_i,a_f\", sort=\"a_f desc, a_i desc\"), over=\"a_f\")");
stream = new UniqueStream(expression, factory);
stream.setStreamContext(streamContext);
tuples = getTuples(stream);
assert (tuples.size() == 4);
assertOrder(tuples, 4, 3, 1, 2);
// Basic w/multi comp
expression = StreamExpressionParser.parse("unique(search(" + COLLECTIONORALIAS + ", q=*:*, fl=\"id,a_s,a_i,a_f\", sort=\"a_f asc, a_i asc\"), over=\"a_f, a_i\")");
stream = new UniqueStream(expression, factory);
stream.setStreamContext(streamContext);
tuples = getTuples(stream);
assert (tuples.size() == 5);
assertOrder(tuples, 0, 2, 1, 3, 4);
// full factory w/multi comp
stream = factory.constructStream("unique(search(" + COLLECTIONORALIAS + ", q=*:*, fl=\"id,a_s,a_i,a_f\", sort=\"a_f asc, a_i asc\"), over=\"a_f, a_i\")");
stream.setStreamContext(streamContext);
tuples = getTuples(stream);
assert (tuples.size() == 5);
assertOrder(tuples, 0, 2, 1, 3, 4);
} finally {
solrClientCache.close();
}
}
use of org.apache.solr.client.solrj.io.Tuple in project lucene-solr by apache.
the class StreamExpressionTest method testConvolution.
@Test
public void testConvolution() throws Exception {
UpdateRequest updateRequest = new UpdateRequest();
int i = 0;
while (i < 50) {
updateRequest.add(id, "id_" + (++i), "test_dt", getDateString("2016", "5", "1"), "price_f", "400.00");
}
while (i < 100) {
updateRequest.add(id, "id_" + (++i), "test_dt", getDateString("2015", "5", "1"), "price_f", "300.0");
}
while (i < 150) {
updateRequest.add(id, "id_" + (++i), "test_dt", getDateString("2014", "5", "1"), "price_f", "500.0");
}
while (i < 250) {
updateRequest.add(id, "id_" + (++i), "test_dt", getDateString("2013", "5", "1"), "price_f", "100.00");
}
updateRequest.commit(cluster.getSolrClient(), COLLECTIONORALIAS);
String expr = "timeseries(" + COLLECTIONORALIAS + ", q=\"*:*\", start=\"2013-01-01T01:00:00.000Z\", " + "end=\"2016-12-01T01:00:00.000Z\", " + "gap=\"+1YEAR\", " + "field=\"test_dt\", " + "count(*), sum(price_f), max(price_f), min(price_f))";
String cexpr = "let(a=" + expr + ", b=select(" + expr + ",mult(2, count(*)) as nvalue), c=col(a, count(*)), d=col(b, nvalue), tuple(colc=c, cold=d, conv=conv(c,d)))";
ModifiableSolrParams paramsLoc = new ModifiableSolrParams();
paramsLoc.set("expr", cexpr);
paramsLoc.set("qt", "/stream");
String url = cluster.getJettySolrRunners().get(0).getBaseUrl().toString() + "/" + COLLECTIONORALIAS;
TupleStream solrStream = new SolrStream(url, paramsLoc);
StreamContext context = new StreamContext();
solrStream.setStreamContext(context);
List<Tuple> tuples = getTuples(solrStream);
assertTrue(tuples.size() == 1);
List<Number> convolution = (List<Number>) (tuples.get(0)).get("conv");
assertTrue(convolution.size() == 7);
assertTrue(convolution.get(0).equals(20000D));
assertTrue(convolution.get(1).equals(20000D));
assertTrue(convolution.get(2).equals(25000D));
assertTrue(convolution.get(3).equals(30000D));
assertTrue(convolution.get(4).equals(15000D));
assertTrue(convolution.get(5).equals(10000D));
assertTrue(convolution.get(6).equals(5000D));
}
use of org.apache.solr.client.solrj.io.Tuple in project lucene-solr by apache.
the class StreamExpressionTest method testSubFacetStream.
@Test
public void testSubFacetStream() throws Exception {
new UpdateRequest().add(id, "0", "level1_s", "hello0", "level2_s", "a", "a_i", "0", "a_f", "1").add(id, "2", "level1_s", "hello0", "level2_s", "a", "a_i", "2", "a_f", "2").add(id, "3", "level1_s", "hello3", "level2_s", "a", "a_i", "3", "a_f", "3").add(id, "4", "level1_s", "hello4", "level2_s", "a", "a_i", "4", "a_f", "4").add(id, "1", "level1_s", "hello0", "level2_s", "b", "a_i", "1", "a_f", "5").add(id, "5", "level1_s", "hello3", "level2_s", "b", "a_i", "10", "a_f", "6").add(id, "6", "level1_s", "hello4", "level2_s", "b", "a_i", "11", "a_f", "7").add(id, "7", "level1_s", "hello3", "level2_s", "b", "a_i", "12", "a_f", "8").add(id, "8", "level1_s", "hello3", "level2_s", "b", "a_i", "13", "a_f", "9").add(id, "9", "level1_s", "hello0", "level2_s", "b", "a_i", "14", "a_f", "10").commit(cluster.getSolrClient(), COLLECTIONORALIAS);
String clause;
TupleStream stream;
List<Tuple> tuples;
StreamFactory factory = new StreamFactory().withCollectionZkHost("collection1", cluster.getZkServer().getZkAddress()).withFunctionName("facet", FacetStream.class).withFunctionName("sum", SumMetric.class).withFunctionName("min", MinMetric.class).withFunctionName("max", MaxMetric.class).withFunctionName("avg", MeanMetric.class).withFunctionName("count", CountMetric.class);
// Basic test
clause = "facet(" + "collection1, " + "q=\"*:*\", " + "buckets=\"level1_s, level2_s\", " + "bucketSorts=\"sum(a_i) desc, sum(a_i) desc)\", " + "bucketSizeLimit=100, " + "sum(a_i), count(*)" + ")";
stream = factory.constructStream(clause);
tuples = getTuples(stream);
assert (tuples.size() == 6);
Tuple tuple = tuples.get(0);
String bucket1 = tuple.getString("level1_s");
String bucket2 = tuple.getString("level2_s");
Double sumi = tuple.getDouble("sum(a_i)");
Double count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello3"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 35);
assertTrue(count.doubleValue() == 3);
tuple = tuples.get(1);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello0"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 15);
assertTrue(count.doubleValue() == 2);
tuple = tuples.get(2);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello4"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 11);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(3);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello4"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 4);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(4);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello3"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 3);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(5);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello0"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 2);
assertTrue(count.doubleValue() == 2);
clause = "facet(" + "collection1, " + "q=\"*:*\", " + "buckets=\"level1_s, level2_s\", " + "bucketSorts=\"level1_s desc, level2_s desc)\", " + "bucketSizeLimit=100, " + "sum(a_i), count(*)" + ")";
stream = factory.constructStream(clause);
tuples = getTuples(stream);
assert (tuples.size() == 6);
tuple = tuples.get(0);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello4"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 11);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(1);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello4"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 4);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(2);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello3"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 35);
assertTrue(count.doubleValue() == 3);
tuple = tuples.get(3);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello3"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 3);
assertTrue(count.doubleValue() == 1);
tuple = tuples.get(4);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello0"));
assertTrue(bucket2.equals("b"));
assertTrue(sumi.longValue() == 15);
assertTrue(count.doubleValue() == 2);
tuple = tuples.get(5);
bucket1 = tuple.getString("level1_s");
bucket2 = tuple.getString("level2_s");
sumi = tuple.getDouble("sum(a_i)");
count = tuple.getDouble("count(*)");
assertTrue(bucket1.equals("hello0"));
assertTrue(bucket2.equals("a"));
assertTrue(sumi.longValue() == 2);
assertTrue(count.doubleValue() == 2);
}
use of org.apache.solr.client.solrj.io.Tuple in project lucene-solr by apache.
the class StreamExpressionTest method testLength.
@Test
public void testLength() throws Exception {
UpdateRequest updateRequest = new UpdateRequest();
updateRequest.add(id, "1", "price_f", "100.0", "col_s", "a", "order_i", "1");
updateRequest.add(id, "2", "price_f", "200.0", "col_s", "a", "order_i", "2");
updateRequest.add(id, "3", "price_f", "300.0", "col_s", "a", "order_i", "3");
updateRequest.add(id, "4", "price_f", "100.0", "col_s", "a", "order_i", "4");
updateRequest.add(id, "5", "price_f", "200.0", "col_s", "a", "order_i", "5");
updateRequest.add(id, "6", "price_f", "400.0", "col_s", "a", "order_i", "6");
updateRequest.add(id, "7", "price_f", "600.0", "col_s", "a", "order_i", "7");
updateRequest.add(id, "8", "price_f", "200.0", "col_s", "b", "order_i", "1");
updateRequest.add(id, "9", "price_f", "400.0", "col_s", "b", "order_i", "2");
updateRequest.add(id, "10", "price_f", "600.0", "col_s", "b", "order_i", "3");
updateRequest.add(id, "11", "price_f", "200.0", "col_s", "b", "order_i", "4");
updateRequest.add(id, "12", "price_f", "400.0", "col_s", "b", "order_i", "5");
updateRequest.add(id, "13", "price_f", "800.0", "col_s", "b", "order_i", "6");
updateRequest.add(id, "14", "price_f", "1200.0", "col_s", "b", "order_i", "7");
updateRequest.commit(cluster.getSolrClient(), COLLECTIONORALIAS);
String expr1 = "search(" + COLLECTIONORALIAS + ", q=\"col_s:a\", fl=\"price_f, order_i\", sort=\"order_i asc\")";
String expr2 = "search(" + COLLECTIONORALIAS + ", q=\"col_s:b\", fl=\"price_f, order_i\", sort=\"order_i asc\")";
String cexpr = "let(a=" + expr1 + ", b=" + expr2 + ", c=col(a, price_f), d=col(b, price_f), e=regress(c, d), tuple(regress=e, p=predict(e, 300), l=length(d)))";
ModifiableSolrParams paramsLoc = new ModifiableSolrParams();
paramsLoc.set("expr", cexpr);
paramsLoc.set("qt", "/stream");
String url = cluster.getJettySolrRunners().get(0).getBaseUrl().toString() + "/" + COLLECTIONORALIAS;
TupleStream solrStream = new SolrStream(url, paramsLoc);
StreamContext context = new StreamContext();
solrStream.setStreamContext(context);
List<Tuple> tuples = getTuples(solrStream);
assertTrue(tuples.size() == 1);
Tuple tuple = tuples.get(0);
Map regression = (Map) tuple.get("regress");
double slope = (double) regression.get("slope");
double intercept = (double) regression.get("intercept");
double length = tuple.getDouble("l");
assertTrue(slope == 2.0D);
assertTrue(intercept == 0.0D);
double prediction = tuple.getDouble("p");
assertTrue(prediction == 600.0D);
assertTrue(length == 7);
}
Aggregations