22286: Amlg lang improvements: removes query_weighted_sample, weighte…

…d_rand, updates rand, query_sample, sort, reduce, MAJOR (#327)

Removes query_weighted_sample and weighted_rand, updates query_sample
and rand to handle functionality.
Removes access to (target) from reduce and sort opcodes for performance
and consistency reasons.
Improves documentation and clarifies some opcode use.

src/Amalgam/Opcodes.cpp

@@ -69,7 +69,6 @@ void StringInternPool::InitializeStaticStrings()
 
 	//simulation and operations
 	EmplaceNodeTypeString(ENT_RAND, "rand");
-	EmplaceNodeTypeString(ENT_WEIGHTED_RAND, "weighted_rand");
 	EmplaceNodeTypeString(ENT_GET_RAND_SEED, "get_rand_seed");
 	EmplaceNodeTypeString(ENT_SET_RAND_SEED, "set_rand_seed");
 	EmplaceNodeTypeString(ENT_SYSTEM_TIME, "system_time");
@@ -257,7 +256,6 @@ void StringInternPool::InitializeStaticStrings()
 	EmplaceNodeTypeString(ENT_COMPUTE_ON_CONTAINED_ENTITIES, "compute_on_contained_entities");
 	EmplaceNodeTypeString(ENT_QUERY_SELECT, "query_select");
 	EmplaceNodeTypeString(ENT_QUERY_SAMPLE, "query_sample");
-	EmplaceNodeTypeString(ENT_QUERY_WEIGHTED_SAMPLE, "query_weighted_sample");
 	EmplaceNodeTypeString(ENT_QUERY_IN_ENTITY_LIST, "query_in_entity_list");
 	EmplaceNodeTypeString(ENT_QUERY_NOT_IN_ENTITY_LIST, "query_not_in_entity_list");
 	EmplaceNodeTypeString(ENT_QUERY_EXISTS, "query_exists");

src/Amalgam/Opcodes.h

@@ -48,7 +48,6 @@ enum EvaluableNodeType : uint8_t
 
 	//simulation and operations
 	ENT_RAND,
-	ENT_WEIGHTED_RAND,
 	ENT_GET_RAND_SEED,
 	ENT_SET_RAND_SEED,
 	ENT_SYSTEM_TIME,
@@ -238,7 +237,6 @@ enum EvaluableNodeType : uint8_t
 	ENT_COMPUTE_ON_CONTAINED_ENTITIES,
 	ENT_QUERY_SELECT,
 	ENT_QUERY_SAMPLE,
-	ENT_QUERY_WEIGHTED_SAMPLE,
 	ENT_QUERY_IN_ENTITY_LIST,
 	ENT_QUERY_NOT_IN_ENTITY_LIST,
 	ENT_QUERY_EXISTS,
@@ -337,7 +335,7 @@ constexpr OrderedChildNodeType GetInstructionOrderedChildNodeType(EvaluableNodeT
 	case ENT_SET_ENTITY_RAND_SEED:
 	case ENT_CREATE_ENTITIES:
 	case ENT_CONTAINED_ENTITIES:	case ENT_COMPUTE_ON_CONTAINED_ENTITIES:
-	case ENT_QUERY_SELECT:			case ENT_QUERY_SAMPLE:									case ENT_QUERY_WEIGHTED_SAMPLE:
+	case ENT_QUERY_SELECT:			case ENT_QUERY_SAMPLE:
 	case ENT_QUERY_IN_ENTITY_LIST:	case ENT_QUERY_NOT_IN_ENTITY_LIST:
 	case ENT_QUERY_EXISTS:			case ENT_QUERY_NOT_EXISTS:
 	case ENT_QUERY_EQUALS:			case ENT_QUERY_NOT_EQUALS:
@@ -375,7 +373,7 @@ constexpr OrderedChildNodeType GetInstructionOrderedChildNodeType(EvaluableNodeT
 	case ENT_GET:
 	case ENT_TARGET:					case ENT_CURRENT_INDEX:		case ENT_CURRENT_VALUE:		case ENT_PREVIOUS_RESULT:
 	case ENT_OPCODE_STACK:				case ENT_STACK:				case ENT_ARGS:
-	case ENT_RAND:						case ENT_WEIGHTED_RAND:		case ENT_GET_RAND_SEED:		case ENT_SET_RAND_SEED:
+	case ENT_RAND:						case ENT_GET_RAND_SEED:		case ENT_SET_RAND_SEED:
 	case ENT_SYSTEM_TIME:
 	case ENT_GET_DIGITS:				case ENT_SET_DIGITS:
 	case ENT_FLOOR:						case ENT_CEILING:			case ENT_ROUND:
@@ -482,7 +480,7 @@ constexpr bool DoesEvaluableNodeTypeCreateScope(EvaluableNodeType t)
 constexpr bool IsEvaluableNodeTypeQuery(EvaluableNodeType t)
 {
 	return (t == ENT_QUERY_SELECT || t == ENT_QUERY_IN_ENTITY_LIST || t == ENT_QUERY_NOT_IN_ENTITY_LIST
-		|| t == ENT_QUERY_SAMPLE || t == ENT_QUERY_WEIGHTED_SAMPLE || t == ENT_QUERY_EXISTS || t == ENT_QUERY_NOT_EXISTS
+		|| t == ENT_QUERY_SAMPLE || t == ENT_QUERY_EXISTS || t == ENT_QUERY_NOT_EXISTS
 		|| t == ENT_QUERY_EQUALS || t == ENT_QUERY_NOT_EQUALS
 		|| t == ENT_QUERY_BETWEEN || t == ENT_QUERY_NOT_BETWEEN || t == ENT_QUERY_AMONG || t == ENT_QUERY_NOT_AMONG
 		|| t == ENT_QUERY_MAX || t == ENT_QUERY_MIN || t == ENT_QUERY_SUM || t == ENT_QUERY_MODE

src/Amalgam/amlg_code/full_test.amlg

@@ -1163,26 +1163,17 @@
 
  (print (rand 50 4) "\n")
 
- (print "--weighted_rand--\n")
- (print (weighted_rand (associate "a" .25 "b" .75)) "\n")
- (print (weighted_rand (associate "a" .25 "b" .75) 4) "\n")
+ (print (rand (associate "a" .25 "b" .75)) "\n")
+ (print (rand (associate "a" .25 "b" .75) 16) "\n")
 
- (print (weighted_rand (list (list "a" "b") (list .25 .75))  ) "\n")
- (print (weighted_rand (list (list "a" "b") (list .25 .75)) 4) "\n")
- (print (weighted_rand (list (list "a" "b") (list 0 0)) 4 ) "\n" )
-
- (print "infinity test c or d: " (weighted_rand (associate "a" .25 "b" .75 "c" .infinity "d" .infinity) 4) "\n")
- (print "infinity test c or d: " (weighted_rand (list (list "a" "b" "c" "d") (list .25 .75 .infinity .infinity)) 4) "\n")
+ (print "infinity test c or d: " (rand (associate "a" .25 "b" .75 "c" .infinity "d" .infinity) 4) "\n")
 
  ;these should come out somewhere near the correct proportions
- (print (zip (lambda (+ (current_value 1) (current_value))) (weighted_rand (associate "a" .25 "b" .5 "c" .25) 100) 1) "\n")
-
-  ;these should come out somewhere near the correct proportions
- (print (zip (lambda (+ (current_value 1) (current_value))) (weighted_rand (list (list "a" "b" "c") (list .25 .5 .25)) 100) 1) "\n")
+ (print (zip (lambda (+ (current_value 1) (current_value))) (rand (associate "a" .25 "b" .5 "c" .25) 100) 1) "\n")
 
  ;these should be weighted toward smaller numbers
  (print
-	(weighted_rand
+	(rand
 		(zip
 			(range 1 10)
 			(map (lambda
@@ -2519,38 +2510,38 @@
 
 
  (print (contained_entities "TestContainerExec" (list
-  (query_sample 1 (rand) )
+  (query_sample 1 (null) (rand) )
  )))
 
  (print (contained_entities "TestContainerExec" (list
-  (query_sample 1 (null) )
+  (query_sample 1 (null) (null) )
  )))
 
- (print "--query_weighted_sample--\n")
+ (print "with weights\n")
 
  (print (contained_entities "TestContainerExec" (list
-   (query_weighted_sample "weight")
+   (query_sample 1 "weight")
   )))
 
  (print (contained_entities "TestContainerExec" (list
-  (query_weighted_sample "weight")
+  (query_sample 1 "weight")
  )))
 
  (print (contained_entities "TestContainerExec" (list
-  (query_weighted_sample "weight" 20 (rand) )
+  (query_sample 20 "weight" (rand) )
  )))
 
  (print (contained_entities "TestContainerExec" (list
-  (query_weighted_sample "weight" 20 (null) )
+  (query_sample 20 "weight" (null) )
  )))
 
  (print (contained_entities "TestContainerExec" (list
   (query_not_in_entity_list (list "Child1"))
-  (query_weighted_sample "weight" 10 (rand) )
+  (query_sample 10 "weight" (rand) )
  )))
 
  (print (contained_entities "TestContainerExec" (list
-  (query_weighted_sample "weight" 10 (rand) )
+  (query_sample 10 "weight" (rand) )
   (query_not_in_entity_list (list "Child1"))
  )))
 

src/Amalgam/amlg_code/test.amlg

@@ -1,14 +1,80 @@
 (seq
- (create_entities "persist_transactions" (list 1 2 3 4))
- (create_entities (list "persist_transactions" "child") (list 5 6 7))
- (create_entities (list "persist_transactions" "child" "doublechild1") (lambda (list ##eight 8 ##nine 9)))
- (create_entities (list "persist_transactions" "child" "doublechild3") (list 12 13))
- (store_entity "amlg_code/test_output/persist_transactions.caml" "persist_transactions" (null) (true) {flatten (true) transactional (true)})
- (create_entities (list "persist_transactions" "child" "doublechild2") (list 10 11))
- (destroy_entities (list "persist_transactions" "child" "doublechild3"))
- (assign_to_entities (list "persist_transactions" "child" "doublechild1") {eight 88})
-
- (load_entity "amlg_code/test_output/persist_transactions.caml" "persist_transactions_copy" (null) (false) {execute_on_load (true) transactional (true) require_version_compatibility (true)})
- (print (difference_entities "persist_transactions_copy" "persist_transactions") "\n")
- (destroy_entities "persist_transactions_copy" "persist_transactions")
-)
+
+
+ (create_entities "TestContainerExec"
+	(lambda (parallel
+    ##^a 3
+	##b (contained_entities)
+	##c (+ x 1)
+	##d (call_entity "Child5" "q" (assoc x x))
+	##!e 12
+	##x 4
+	##y 5
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child1")
+	(lambda (parallel
+	##x 3
+	##y 4
+	##!e 7
+	##weight 0.45
+	##weight_eq 1
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child2")
+	(lambda (parallel
+	##x -1
+	##y -1
+	##weight 0.45
+	##weight_eq 1
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child3")
+	(lambda (parallel
+	##x 100
+	##y 100
+	##weight 0.02
+	##weight_eq 1
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child4")
+	(lambda (parallel
+	##x 100
+	##y 100
+	##radius 400
+	##weight 0.02
+	##weight_eq 1
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child5")
+	(lambda (parallel
+	##p 3
+	##q (+ x (call_container "^a"))
+	##bar "crunchy"
+	##weight 0.02
+	##weight_eq 1
+	))
+	)
+ (create_entities (list "TestContainerExec" "Child6")
+	(lambda (parallel
+	##x 1
+	##y 2
+	##bar "not crunchy"
+	##weight 0.02
+	##weight_eq 1
+	))
+ )
+ (create_entities (list "TestContainerExec" "Child7")
+	(lambda (parallel
+	##x 0
+	##y 10
+	##weight 0.02
+	##weight_eq 1
+	))
+ )
+
+(print (contained_entities "TestContainerExec" (list
+   (query_sample 1 "weight")
+  )))
+
+ )

src/Amalgam/entity/EntityQueries.cpp

@@ -19,7 +19,6 @@ bool EntityQueryCondition::DoesEntityMatchCondition(Entity *e)
 
 		case ENT_QUERY_SELECT:
 		case ENT_QUERY_SAMPLE:
-		case ENT_QUERY_WEIGHTED_SAMPLE:
 			//it does not fail the condition here - needs to be checked elsewhere
 			return true;
 
@@ -352,75 +351,52 @@ EvaluableNodeReference EntityQueryCondition::GetMatchingEntities(Entity *contain
 		else //just use a random seed
 			random_stream.SetState("12345");
 
-		//select num_to_select entities and save them in the sample vector
-		for(size_t i = 0; i < num_to_sample; i++)
+		if(singleLabel == StringInternPool::NOT_A_STRING_ID)
 		{
-			size_t index_to_swap = randomStream.RandSize(num_entities);
-			Entity *selected = matching_entities[index_to_swap];
-			samples.emplace_back(selected);
+			//select num_to_select entities and save them in the sample vector
+			for(size_t i = 0; i < num_to_sample; i++)
+			{
+				size_t index_to_swap = randomStream.RandSize(num_entities);
+				Entity *selected = matching_entities[index_to_swap];
+				samples.emplace_back(selected);
+			}
 		}
-
-		//swap samples vector with the matching_entities 
-		std::swap(matching_entities, samples);
-		return EvaluableNodeReference::Null();
-	}
-
-	case ENT_QUERY_WEIGHTED_SAMPLE:
-	{
-		size_t num_entities = matching_entities.size();
-		size_t num_to_sample = static_cast<size_t>(maxToRetrieve);
-		auto weight_label_id = singleLabel;
-
-		if(num_entities == 0 || num_to_sample == 0)
+		else //weighted
 		{
-			matching_entities.clear();
-			return EvaluableNodeReference::Null();
-		}
-
-		//retrieve weights
-		std::vector<double> entity_weights;
-		entity_weights.reserve(num_to_sample);
+			//retrieve weights
+			std::vector<double> entity_weights;
+			entity_weights.reserve(num_to_sample);
 
-		//retrieve and accumulate weights
-		for(size_t i = 0; i < matching_entities.size(); i++)
-		{
-			double value;
-			Entity *e = matching_entities[i];
-			if(e != nullptr && e->GetValueAtLabelAsNumber(weight_label_id, value))
+			//retrieve and accumulate weights
+			for(Entity *e : matching_entities)
 			{
-				if(FastIsNaN(value))
-					value = 0.0;
+				double value;
+				if(e != nullptr && e->GetValueAtLabelAsNumber(singleLabel, value))
+				{
+					if(FastIsNaN(value))
+						value = 0.0;
 
-				entity_weights.push_back(value);
+					entity_weights.push_back(value);
+				}
+				else
+				{
+					entity_weights.push_back(0.0);
+				}
 			}
-			else
+
+			//if just one sample, brute-force it
+			if(num_to_sample == 1)
 			{
-				entity_weights.push_back(0.0);
+				size_t selected_index = WeightedDiscreteRandomSample(entity_weights, random_stream, true);
+				Entity *selected = matching_entities[selected_index];
+				samples.emplace_back(selected);
+			}
+			else //build temporary cache and query
+			{
+				WeightedDiscreteRandomStreamTransform<Entity *> wdrst(matching_entities, entity_weights, true);
+				for(size_t i = 0; i < num_to_sample; i++)
+					samples.emplace_back(wdrst.WeightedDiscreteRand(random_stream));
 			}
-		}
-
-		//obtain random stream either from the condition or use a default one
-		RandomStream random_stream;
-		if(hasRandomStream)
-			random_stream = randomStream.CreateOtherStreamViaRand();
-		else //just use a random seed
-			random_stream.SetState("12345");
-
-		std::vector<Entity *> samples;
-		samples.reserve(num_to_sample);
-
-		//if just one sample, brute-force it
-		if(num_to_sample == 1)
-		{
-			size_t selected_index = WeightedDiscreteRandomSample(entity_weights, random_stream, true);
-			Entity *selected = matching_entities[selected_index];
-			samples.emplace_back(selected);
-		}
-		else //build temporary cache and query
-		{
-			WeightedDiscreteRandomStreamTransform<Entity *> wdrst(matching_entities, entity_weights, true);
-			for(size_t i = 0; i < num_to_sample; i++)
-				samples.emplace_back(wdrst.WeightedDiscreteRand(random_stream));
 		}
 
 		//swap samples vector with the matching_entities 

src/Amalgam/entity/EntityQueryBuilder.h

@@ -652,12 +652,12 @@ namespace EntityQueryBuilder
 		{
 			case ENT_QUERY_SELECT:
 			{
-				cur_condition->maxToRetrieve = (ocn.size() >= 1) ? EvaluableNode::ToNumber(ocn[0], 0.0) : 0;
+				cur_condition->maxToRetrieve = (ocn.size() > 0) ? EvaluableNode::ToNumber(ocn[0], 0.0) : 0;
 
-				cur_condition->hasStartOffset = (ocn.size() >= 2);
+				cur_condition->hasStartOffset = (ocn.size() > 1);
 				cur_condition->startOffset = cur_condition->hasStartOffset ? static_cast<size_t>(EvaluableNode::ToNumber(ocn[1], 0.0)) : 0;
 
-				cur_condition->hasRandomStream = (ocn.size() >= 3 && !EvaluableNode::IsNull(ocn[2]));
+				cur_condition->hasRandomStream = (ocn.size() > 2 && !EvaluableNode::IsNull(ocn[2]));
 				if(cur_condition->hasRandomStream)
 					cur_condition->randomStream.SetState(EvaluableNode::ToString(ocn[2]));
 				else
@@ -668,23 +668,14 @@ namespace EntityQueryBuilder
 			case ENT_QUERY_SAMPLE:
 			{
 				cur_condition->maxToRetrieve = (ocn.size() > 0) ? EvaluableNode::ToNumber(ocn[0], 0.0) : 1;
-				cur_condition->hasRandomStream = (ocn.size() > 1 && !EvaluableNode::IsNull(ocn[1]));
-				if(cur_condition->hasRandomStream)
-					cur_condition->randomStream.SetState(EvaluableNode::ToString(ocn[1]));
-				else
-					cur_condition->randomStream = rs.CreateOtherStreamViaRand();
-			    break;
-			}
-			case ENT_QUERY_WEIGHTED_SAMPLE:
-			{
-				cur_condition->singleLabel = (ocn.size() > 0) ? EvaluableNode::ToStringIDIfExists(ocn[0]) : StringInternPool::NOT_A_STRING_ID;
-				cur_condition->maxToRetrieve = (ocn.size() > 1) ? EvaluableNode::ToNumber(ocn[1], 0.0) : 1;
+				cur_condition->singleLabel = (ocn.size() > 1) ? EvaluableNode::ToStringIDIfExists(ocn[1]) : StringInternPool::NOT_A_STRING_ID;
+
 				cur_condition->hasRandomStream = (ocn.size() > 2 && !EvaluableNode::IsNull(ocn[2]));
 				if(cur_condition->hasRandomStream)
 					cur_condition->randomStream.SetState(EvaluableNode::ToString(ocn[2]));
 				else
 					cur_condition->randomStream = rs.CreateOtherStreamViaRand();
-				break;
+			    break;
 			}
 			case ENT_QUERY_IN_ENTITY_LIST:
 			case ENT_QUERY_NOT_IN_ENTITY_LIST: