Revert "Revert "Require constant() for large constants and fix a type resolving bug""

30 files changed, 556 insertions, 74 deletions

diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/FeatureNames.java b/config-model/src/main/java/com/yahoo/searchdefinition/FeatureNames.java
index 2f41b172ab6..1e133d0b8f4 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/FeatureNames.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/FeatureNames.java
@@ -34,6 +34,12 @@ public class FeatureNames {
         return name.equals("attribute") || name.equals("constant") || name.equals("query");
     }
 
+    /** Returns true if this is a constant */
+    public static boolean isConstantFeature(Reference reference) {
+        if ( ! isSimpleFeature(reference)) return false;
+        return reference.name().equals("constant");
+    }
+
     /**
      * Returns the single argument of the given feature name, without any quotes,
      * or empty if it is not a valid query, attribute or constant feature name
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/RankProfile.java b/config-model/src/main/java/com/yahoo/searchdefinition/RankProfile.java
index b3853b36aa5..d738929f721 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/RankProfile.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/RankProfile.java
@@ -756,7 +756,7 @@ public class RankProfile implements Serializable, Cloneable {
             for (FieldDescription field : queryProfileType.declaredFields().values()) {
                 TensorType type = field.getType().asTensorType();
                 Optional<Reference> feature = Reference.simple(field.getName());
-                if ( ! feature.isPresent() || ! feature.get().name().equals("query")) continue;
+                if ( feature.isEmpty() || ! feature.get().name().equals("query")) continue;
 
                 TensorType existingType = context.getType(feature.get());
                 if ( ! Objects.equals(existingType, context.defaultTypeOf(feature.get())))
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/SearchBuilder.java b/config-model/src/main/java/com/yahoo/searchdefinition/SearchBuilder.java
index 9804b0b6329..a84db895b02 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/SearchBuilder.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/SearchBuilder.java
@@ -8,7 +8,9 @@ import com.yahoo.config.model.test.MockApplicationPackage;
 import com.yahoo.document.DocumentTypeManager;
 import com.yahoo.io.IOUtils;
 import com.yahoo.io.reader.NamedReader;
+import com.yahoo.search.query.profile.QueryProfile;
 import com.yahoo.search.query.profile.QueryProfileRegistry;
+import com.yahoo.search.query.profile.config.QueryProfileXMLReader;
 import com.yahoo.searchdefinition.derived.SearchOrderer;
 import com.yahoo.searchdefinition.document.SDDocumentType;
 import com.yahoo.searchdefinition.parser.ParseException;
@@ -394,14 +396,13 @@ public class SearchBuilder {
     }
 
     public static SearchBuilder createFromDirectory(String dir) throws IOException, ParseException {
-        return createFromDirectory(dir, new RankProfileRegistry(), new QueryProfileRegistry());
+        return createFromDirectory(dir, new RankProfileRegistry());
     }
     public static SearchBuilder createFromDirectory(String dir,
-                                                    RankProfileRegistry rankProfileRegistry,
-                                                    QueryProfileRegistry queryProfileRegistry) throws IOException, ParseException {
+                                                    RankProfileRegistry rankProfileRegistry) throws IOException, ParseException {
         SearchBuilder builder = new SearchBuilder(MockApplicationPackage.fromSearchDefinitionDirectory(dir),
                                                   rankProfileRegistry,
-                                                  queryProfileRegistry);
+                                                  createQueryProfileRegistryFromDirectory(dir));
         for (Iterator<Path> i = Files.list(new File(dir).toPath()).filter(p -> p.getFileName().toString().endsWith(".sd")).iterator(); i.hasNext(); ) {
             builder.importFile(i.next());
         }
@@ -409,6 +410,12 @@ public class SearchBuilder {
         return builder;
     }
 
+    private static QueryProfileRegistry createQueryProfileRegistryFromDirectory(String dir) {
+        File queryProfilesDir = new File(dir, "query-profiles");
+        if ( ! queryProfilesDir.exists()) return new QueryProfileRegistry();
+        return new QueryProfileXMLReader().read(queryProfilesDir.toString());
+    }
+
     // TODO: The build methods below just call the create methods above - remove
 
     /**
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ConstantTensorTransformer.java b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ConstantTensorTransformer.java
index caf5f0442eb..6991e2b978b 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ConstantTensorTransformer.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ConstantTensorTransformer.java
@@ -49,13 +49,13 @@ public class ConstantTensorTransformer extends ExpressionTransformer<RankProfile
     }
 
     private ExpressionNode transformConstantReference(ReferenceNode node, RankProfileTransformContext context) {
-        Reference constantReference = FeatureNames.asConstantFeature(node.getName());
+        Reference constantReference = node.reference();
+        if ( ! FeatureNames.isConstantFeature(constantReference) && constantReference.isIdentifier())
+            constantReference = FeatureNames.asConstantFeature(node.getName());
+
         Value value = context.constants().get(node.getName());
-        if (value == null || value.type().rank() == 0) {
-            if (context.rankProfile().rankingConstants().get(node.getName()) != null) // Large constants: Transform reference but don't add value
-                return new ReferenceNode(constantReference);
-            return node;
-        }
+        if (value == null || value.type().rank() == 0) return node;
+
         TensorValue tensorValue = (TensorValue)value;
         String tensorType = tensorValue.asTensor().type().toString();
         context.rankProperties().put(constantReference.toString() + ".value", tensorValue.toString());
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ExpressionTransforms.java b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ExpressionTransforms.java
index cbabfffb7a1..6fdf448a39b 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ExpressionTransforms.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/ExpressionTransforms.java
@@ -6,6 +6,7 @@ import com.yahoo.searchlib.rankingexpression.RankingExpression;
 import com.yahoo.searchlib.rankingexpression.transform.ConstantDereferencer;
 import com.yahoo.searchlib.rankingexpression.transform.ExpressionTransformer;
 import com.yahoo.searchlib.rankingexpression.transform.Simplifier;
+import com.yahoo.searchlib.rankingexpression.transform.TensorMaxMinTransformer;
 
 import java.util.List;
 
@@ -30,7 +31,7 @@ public class ExpressionTransforms {
                                  new ConstantTensorTransformer(),
                                  new FunctionInliner(),
                                  new FunctionShadower(),
-                                 new TensorTransformer(),
+                                 new TensorMaxMinTransformer(),
                                  new Simplifier());
     }
 
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/RankProfileTransformContext.java b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/RankProfileTransformContext.java
index 2c0e1eaa56a..630c8644eb1 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/RankProfileTransformContext.java
+++ b/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/RankProfileTransformContext.java
@@ -24,19 +24,17 @@ public class RankProfileTransformContext extends TransformContext {
     private final ImportedMlModels importedModels;
     private final Map<String, RankProfile.RankingExpressionFunction> inlineFunctions;
     private final Map<String, String> rankProperties = new HashMap<>();
-    private final MapEvaluationTypeContext types;
 
     public RankProfileTransformContext(RankProfile rankProfile,
                                        QueryProfileRegistry queryProfiles,
                                        ImportedMlModels importedModels,
                                        Map<String, Value> constants,
                                        Map<String, RankProfile.RankingExpressionFunction> inlineFunctions) {
-        super(constants);
+        super(constants, rankProfile.typeContext(queryProfiles));
         this.rankProfile = rankProfile;
         this.queryProfiles = queryProfiles;
         this.importedModels = importedModels;
         this.inlineFunctions = inlineFunctions;
-        this.types = rankProfile.typeContext(queryProfiles);
     }
 
     public RankProfile rankProfile() { return rankProfile; }
@@ -45,10 +43,4 @@ public class RankProfileTransformContext extends TransformContext {
     public Map<String, RankProfile.RankingExpressionFunction> inlineFunctions() { return inlineFunctions; }
     public Map<String, String> rankProperties() { return rankProperties; }
 
-    /**
-     * Returns the types known in this context. We may have type information for references
-     * for which no value is available
-     */
-    public MapEvaluationTypeContext types() { return types; }
-
 }
diff --git a/config-model/src/test/derived/neuralnet/neuralnet.sd b/config-model/src/test/derived/neuralnet/neuralnet.sd
new file mode 100644
index 00000000000..f916b35cb75
--- /dev/null
+++ b/config-model/src/test/derived/neuralnet/neuralnet.sd
@@ -0,0 +1,238 @@
+search neuralnet {
+
+    document neuralnet {
+
+        field pinned type int {
+            indexing: attribute
+        }
+
+        field createdAt type long {
+            indexing: attribute
+        }
+
+        field updatedAt type long {
+            indexing: attribute
+        }
+
+        field uvCount type int {
+            indexing: attribute
+        }
+
+        field dvCount type int {
+            indexing: attribute
+        }
+
+        field aVoteCount type int {
+            indexing: attribute
+        }
+
+        field rCount type int {
+            indexing: attribute
+        }
+
+        field uniqueRACount type int {
+            indexing: attribute
+        }
+
+        field rTo type string {
+            indexing: attribute
+        }
+
+        field markedAsAAt type long {
+            indexing: attribute
+        }
+
+        field normalizedTextScore type float {
+            indexing: attribute
+        }
+
+        field t type float {
+            indexing: attribute
+        }
+
+        field relevance type float {
+            indexing: attribute
+        }
+
+        field normalizedCS type float {
+            indexing: attribute
+        }
+
+        field laAt type long {
+            indexing: attribute
+        }
+
+        field hsScore type double {
+            indexing: attribute
+        }
+
+    }
+
+    rank-profile defaultRankProfile inherits default {
+
+        constants {
+            maxSignedSixtyFourBitInteger: 9223372036854775807
+        }
+
+        macro log10_1p(x) {
+            expression: log10(x+1)
+        }
+
+        macro textScoreToUse() {
+            expression: if(isNan(attribute(normalizedTextScore)) == 1, 0, attribute(normalizedTextScore))
+        }
+
+        macro rCountToUse() {
+            expression: if(isNan(attribute(rCount)) == 1, 0, if(attribute(rCount) < 0, 0, attribute(rCount)))
+        }
+
+        macro uniqueRCountToUse() {
+            expression: if(isNan(attribute(uniqueRCount)) == 1, 0, if(attribute(uniqueRACount) < 0, 0, attribute(uniqueRACount)))
+        }
+
+        macro uvCountToUse() {
+            expression: if(isNan(attribute(uvCount)) == 1, 0, if(attribute(uvCount) < 0, 0, attribute(uvCount)))
+        }
+
+        macro dvCountToUse() {
+            expression: if(isNan(attribute(dvCount)) == 1, 0, if(attribute(dvCount) < 0, 0, attribute(dvCount)))
+        }
+
+        macro aVoteCountToUse() {
+            expression: if(isNan(attribute(aVoteCount)) == 1, 0, if(attribute(aVoteCount) < 0, 0, attribute(aVoteCount)))
+        }
+
+        macro totalPR() {
+            expression: uniqueRCountToUse + query(voteToRRatio) * (uvCountToUse - dvCountToUse) - aVoteCountToUse
+        }
+
+        macro totalvote() {
+            expression: query(reportaweight) * aVoteCountToUse + dvCountToUse + query(rweight) * uniqueRCountToUse + uvCountToUse
+        }
+
+        macro phat() {
+            expression: if (totalvote == 0, 0, ( query(rweight) * uniqueRCountToUse + uvCountToUse) / totalvote)
+        }
+
+        macro nCScoreToUse() {
+            expression: if (totalPR > 0, log10(totalPR), 0)
+        }
+
+        macro hsScoreToUse() {
+            expression: attribute(hsScore)
+        }
+
+        macro tScoreToUse() {
+            expression: if (isNan(attribute(t)) == 1, 0.6, attribute(t))
+        }
+
+        macro relevanceScoreToUse() {
+            expression: if (isNan(attribute(relevance)) == 1, 0.254, attribute(relevance))
+        }
+
+        macro freshnessToUse() {
+            expression: if (freshness(createdAt).logscale < 0.01, 0.01, freshness(createdAt).logscale)
+        }
+
+        macro rankedAt() {
+            expression: now
+        }
+
+        macro createdAtToUse() {
+            expression: if(isNan(attribute(createdAt)) == 1, rankedAt, attribute(createdAt))
+        }
+
+        macro laAtToUse() {
+            expression: if(isNan(attribute(laAt)) == 1, attribute(createdAt), attribute(laAt))
+        }
+
+        macro markedAsAAtToUse() {
+            expression: if(isNan(attribute(markedAsAAt)) == 1, maxSignedSixtyFourBitInteger, attribute(markedAsAAt))
+        }
+
+        macro tdToUse() {
+            expression: pow(2, 0 - ((rankedAt - createdAtToUse) / query(decay)))
+        }
+
+        macro commentOverallScore() {
+            expression: query(textweight) * textScoreToUse  + query(communityweight) * nCScoreToUse
+        }
+
+        macro pinScore() {
+            expression: if(isNan(attribute(pinned)) == 1, 0, query(pinweight) * attribute(pinned))
+        }
+
+        macro freshnessRank() {
+            expression: nativeRank + freshness(createdAt)
+        }
+
+        first-phase {
+            expression: nativeRank
+        }
+
+    }
+
+    rank-profile neuralNetworkProfile inherits defaultRankProfile {
+        macro nn_input() {
+            expression {
+                concat(log10_1p(aVoteCountToUse),
+                concat(log10_1p(dvCountToUse),
+                concat(log10_1p(uniqueRCountToUse),
+                concat(log10_1p(uvCountToUse),
+                concat(phat,
+                concat(log10_1p(totalvote),
+                concat(hsScoreToUse,
+                concat(tdToUse,
+                tScoreToUse, x), x), x), x), x), x), x), x)
+            }
+        }
+
+        macro get_model_weights(field) {
+            expression: if(query(field) == 0, constant(field), query(field))
+        }
+
+        macro layer_0() {
+            expression: elu(xw_plus_b(nn_input, get_model_weights(W_0), get_model_weights(b_0), x))
+        }
+        macro layer_1() {
+            expression: elu(xw_plus_b(layer_0, get_model_weights(W_1), get_model_weights(b_1), hidden))
+        }
+        macro layer_out() {
+            expression: sum(xw_plus_b(layer_1, get_model_weights(W_out), get_model_weights(b_out), out))
+        }
+        first-phase {
+            expression: freshnessRank
+        }
+        second-phase {
+            expression: layer_out
+            rerank-count: 2000
+        }
+
+    }
+
+    constant W_0 {
+        file: neural-network-201805/W_0.json
+        type: tensor(x[9],hidden[9])
+    }
+    constant b_0 {
+        file: neural-network-201805/b_0.json
+        type: tensor(hidden[9])
+    }
+    constant W_1 {
+        file: neural-network-201805/W_1.json
+        type: tensor(hidden[9],out[9])
+    }
+    constant b_1 {
+        file: neural-network-201805/b_1.json
+        type: tensor(out[9])
+    }
+    constant W_out {
+        file: neural-network-201805/W_out.json
+        type: tensor(out[9])
+    }
+    constant b_out {
+        file: neural-network-201805/b_out.json
+        type: tensor(out[1])
+    }
+
+}
\ No newline at end of file
diff --git a/config-model/src/test/derived/neuralnet/query-profiles/default.xml b/config-model/src/test/derived/neuralnet/query-profiles/default.xml
new file mode 100644
index 00000000000..eef1aaa7f53
--- /dev/null
+++ b/config-model/src/test/derived/neuralnet/query-profiles/default.xml
@@ -0,0 +1,2 @@
+<query-profile id="default" type="DefaultQueryProfileType">
+</query-profile>
diff --git a/config-model/src/test/derived/neuralnet/query-profiles/types/DefaultQueryProfileType.xml b/config-model/src/test/derived/neuralnet/query-profiles/types/DefaultQueryProfileType.xml
new file mode 100644
index 00000000000..e1659479135
--- /dev/null
+++ b/config-model/src/test/derived/neuralnet/query-profiles/types/DefaultQueryProfileType.xml
@@ -0,0 +1,8 @@
+<query-profile-type id="DefaultQueryProfileType">
+    <field name="ranking.features.query(W_0)" type="tensor(x[9],hidden[9])" />
+    <field name="ranking.features.query(b_0)" type="tensor(hidden[9])" />
+    <field name="ranking.features.query(W_1)" type="tensor(hidden[9],out[9])" />
+    <field name="ranking.features.query(b_1)" type="tensor(out[9])" />
+    <field name="ranking.features.query(W_out)" type="tensor(out[9])" />
+    <field name="ranking.features.query(b_out)" type="tensor(out[1])" />
+</query-profile-type>
diff --git a/config-model/src/test/derived/neuralnet/rank-profiles.cfg b/config-model/src/test/derived/neuralnet/rank-profiles.cfg
new file mode 100644
index 00000000000..4530bff2e20
--- /dev/null
+++ b/config-model/src/test/derived/neuralnet/rank-profiles.cfg
@@ -0,0 +1,198 @@
+rankprofile[].name "default"
+rankprofile[].fef.property[].name "vespa.type.query.b_out"
+rankprofile[].fef.property[].value "tensor(out[1])"
+rankprofile[].fef.property[].name "vespa.type.query.W_out"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_0"
+rankprofile[].fef.property[].value "tensor(hidden[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_1"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_1"
+rankprofile[].fef.property[].value "tensor(hidden[9],out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_0"
+rankprofile[].fef.property[].value "tensor(hidden[9],x[9])"
+rankprofile[].name "unranked"
+rankprofile[].fef.property[].name "vespa.rank.firstphase"
+rankprofile[].fef.property[].value "value(0)"
+rankprofile[].fef.property[].name "vespa.hitcollector.heapsize"
+rankprofile[].fef.property[].value "0"
+rankprofile[].fef.property[].name "vespa.hitcollector.arraysize"
+rankprofile[].fef.property[].value "0"
+rankprofile[].fef.property[].name "vespa.dump.ignoredefaultfeatures"
+rankprofile[].fef.property[].value "true"
+rankprofile[].fef.property[].name "vespa.type.query.b_out"
+rankprofile[].fef.property[].value "tensor(out[1])"
+rankprofile[].fef.property[].name "vespa.type.query.W_out"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_0"
+rankprofile[].fef.property[].value "tensor(hidden[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_1"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_1"
+rankprofile[].fef.property[].value "tensor(hidden[9],out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_0"
+rankprofile[].fef.property[].value "tensor(hidden[9],x[9])"
+rankprofile[].name "defaultRankProfile"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p).rankingScript"
+rankprofile[].fef.property[].value "log10(x + 1)"
+rankprofile[].fef.property[].name "rankingExpression(textScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(normalizedTextScore)) == 1, 0, attribute(normalizedTextScore))"
+rankprofile[].fef.property[].name "rankingExpression(rCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(rCount)) == 1, 0, if (attribute(rCount) < 0, 0, attribute(rCount)))"
+rankprofile[].fef.property[].name "rankingExpression(uniqueRCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(uniqueRCount)) == 1, 0, if (attribute(uniqueRACount) < 0, 0, attribute(uniqueRACount)))"
+rankprofile[].fef.property[].name "rankingExpression(uvCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(uvCount)) == 1, 0, if (attribute(uvCount) < 0, 0, attribute(uvCount)))"
+rankprofile[].fef.property[].name "rankingExpression(dvCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(dvCount)) == 1, 0, if (attribute(dvCount) < 0, 0, attribute(dvCount)))"
+rankprofile[].fef.property[].name "rankingExpression(aVoteCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(aVoteCount)) == 1, 0, if (attribute(aVoteCount) < 0, 0, attribute(aVoteCount)))"
+rankprofile[].fef.property[].name "rankingExpression(totalPR).rankingScript"
+rankprofile[].fef.property[].value "rankingExpression(uniqueRCountToUse) + query(voteToRRatio) * (rankingExpression(uvCountToUse) - rankingExpression(dvCountToUse)) - rankingExpression(aVoteCountToUse)"
+rankprofile[].fef.property[].name "rankingExpression(totalvote).rankingScript"
+rankprofile[].fef.property[].value "query(reportaweight) * rankingExpression(aVoteCountToUse) + rankingExpression(dvCountToUse) + query(rweight) * rankingExpression(uniqueRCountToUse) + rankingExpression(uvCountToUse)"
+rankprofile[].fef.property[].name "rankingExpression(phat).rankingScript"
+rankprofile[].fef.property[].value "if (rankingExpression(totalvote) == 0, 0, (query(rweight) * rankingExpression(uniqueRCountToUse) + rankingExpression(uvCountToUse)) / rankingExpression(totalvote))"
+rankprofile[].fef.property[].name "rankingExpression(nCScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (rankingExpression(totalPR) > 0, log10(rankingExpression(totalPR)), 0)"
+rankprofile[].fef.property[].name "rankingExpression(hsScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "attribute(hsScore)"
+rankprofile[].fef.property[].name "rankingExpression(tScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(t)) == 1, 0.6, attribute(t))"
+rankprofile[].fef.property[].name "rankingExpression(relevanceScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(relevance)) == 1, 0.254, attribute(relevance))"
+rankprofile[].fef.property[].name "rankingExpression(freshnessToUse).rankingScript"
+rankprofile[].fef.property[].value "if (freshness(createdAt).logscale < 0.01, 0.01, freshness(createdAt).logscale)"
+rankprofile[].fef.property[].name "rankingExpression(rankedAt).rankingScript"
+rankprofile[].fef.property[].value "now"
+rankprofile[].fef.property[].name "rankingExpression(createdAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(createdAt)) == 1, rankingExpression(rankedAt), attribute(createdAt))"
+rankprofile[].fef.property[].name "rankingExpression(laAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(laAt)) == 1, attribute(createdAt), attribute(laAt))"
+rankprofile[].fef.property[].name "rankingExpression(markedAsAAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(markedAsAAt)) == 1, 9.223372036854776E18, attribute(markedAsAAt))"
+rankprofile[].fef.property[].name "rankingExpression(tdToUse).rankingScript"
+rankprofile[].fef.property[].value "pow(2,0 - ((rankingExpression(rankedAt) - rankingExpression(createdAtToUse)) / query(decay)))"
+rankprofile[].fef.property[].name "rankingExpression(commentOverallScore).rankingScript"
+rankprofile[].fef.property[].value "query(textweight) * rankingExpression(textScoreToUse) + query(communityweight) * rankingExpression(nCScoreToUse)"
+rankprofile[].fef.property[].name "rankingExpression(pinScore).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(pinned)) == 1, 0, query(pinweight) * attribute(pinned))"
+rankprofile[].fef.property[].name "rankingExpression(freshnessRank).rankingScript"
+rankprofile[].fef.property[].value "nativeRank + freshness(createdAt)"
+rankprofile[].fef.property[].name "vespa.rank.firstphase"
+rankprofile[].fef.property[].value "nativeRank"
+rankprofile[].fef.property[].name "vespa.type.query.b_out"
+rankprofile[].fef.property[].value "tensor(out[1])"
+rankprofile[].fef.property[].name "vespa.type.query.W_out"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_0"
+rankprofile[].fef.property[].value "tensor(hidden[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_1"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_1"
+rankprofile[].fef.property[].value "tensor(hidden[9],out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_0"
+rankprofile[].fef.property[].value "tensor(hidden[9],x[9])"
+rankprofile[].name "neuralNetworkProfile"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p).rankingScript"
+rankprofile[].fef.property[].value "log10(x + 1)"
+rankprofile[].fef.property[].name "rankingExpression(textScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(normalizedTextScore)) == 1, 0, attribute(normalizedTextScore))"
+rankprofile[].fef.property[].name "rankingExpression(rCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(rCount)) == 1, 0, if (attribute(rCount) < 0, 0, attribute(rCount)))"
+rankprofile[].fef.property[].name "rankingExpression(uniqueRCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(uniqueRCount)) == 1, 0, if (attribute(uniqueRACount) < 0, 0, attribute(uniqueRACount)))"
+rankprofile[].fef.property[].name "rankingExpression(uvCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(uvCount)) == 1, 0, if (attribute(uvCount) < 0, 0, attribute(uvCount)))"
+rankprofile[].fef.property[].name "rankingExpression(dvCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(dvCount)) == 1, 0, if (attribute(dvCount) < 0, 0, attribute(dvCount)))"
+rankprofile[].fef.property[].name "rankingExpression(aVoteCountToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(aVoteCount)) == 1, 0, if (attribute(aVoteCount) < 0, 0, attribute(aVoteCount)))"
+rankprofile[].fef.property[].name "rankingExpression(totalPR).rankingScript"
+rankprofile[].fef.property[].value "rankingExpression(uniqueRCountToUse) + query(voteToRRatio) * (rankingExpression(uvCountToUse) - rankingExpression(dvCountToUse)) - rankingExpression(aVoteCountToUse)"
+rankprofile[].fef.property[].name "rankingExpression(totalvote).rankingScript"
+rankprofile[].fef.property[].value "query(reportaweight) * rankingExpression(aVoteCountToUse) + rankingExpression(dvCountToUse) + query(rweight) * rankingExpression(uniqueRCountToUse) + rankingExpression(uvCountToUse)"
+rankprofile[].fef.property[].name "rankingExpression(phat).rankingScript"
+rankprofile[].fef.property[].value "if (rankingExpression(totalvote) == 0, 0, (query(rweight) * rankingExpression(uniqueRCountToUse) + rankingExpression(uvCountToUse)) / rankingExpression(totalvote))"
+rankprofile[].fef.property[].name "rankingExpression(nCScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (rankingExpression(totalPR) > 0, log10(rankingExpression(totalPR)), 0)"
+rankprofile[].fef.property[].name "rankingExpression(hsScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "attribute(hsScore)"
+rankprofile[].fef.property[].name "rankingExpression(tScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(t)) == 1, 0.6, attribute(t))"
+rankprofile[].fef.property[].name "rankingExpression(relevanceScoreToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(relevance)) == 1, 0.254, attribute(relevance))"
+rankprofile[].fef.property[].name "rankingExpression(freshnessToUse).rankingScript"
+rankprofile[].fef.property[].value "if (freshness(createdAt).logscale < 0.01, 0.01, freshness(createdAt).logscale)"
+rankprofile[].fef.property[].name "rankingExpression(rankedAt).rankingScript"
+rankprofile[].fef.property[].value "now"
+rankprofile[].fef.property[].name "rankingExpression(createdAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(createdAt)) == 1, rankingExpression(rankedAt), attribute(createdAt))"
+rankprofile[].fef.property[].name "rankingExpression(laAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(laAt)) == 1, attribute(createdAt), attribute(laAt))"
+rankprofile[].fef.property[].name "rankingExpression(markedAsAAtToUse).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(markedAsAAt)) == 1, 9.223372036854776E18, attribute(markedAsAAt))"
+rankprofile[].fef.property[].name "rankingExpression(tdToUse).rankingScript"
+rankprofile[].fef.property[].value "pow(2,0 - ((rankingExpression(rankedAt) - rankingExpression(createdAtToUse)) / query(decay)))"
+rankprofile[].fef.property[].name "rankingExpression(commentOverallScore).rankingScript"
+rankprofile[].fef.property[].value "query(textweight) * rankingExpression(textScoreToUse) + query(communityweight) * rankingExpression(nCScoreToUse)"
+rankprofile[].fef.property[].name "rankingExpression(pinScore).rankingScript"
+rankprofile[].fef.property[].value "if (isNan(attribute(pinned)) == 1, 0, query(pinweight) * attribute(pinned))"
+rankprofile[].fef.property[].name "rankingExpression(freshnessRank).rankingScript"
+rankprofile[].fef.property[].value "nativeRank + freshness(createdAt)"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p@af9a8c53ba738798).rankingScript"
+rankprofile[].fef.property[].value "log10(rankingExpression(aVoteCountToUse) + 1)"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p@6ad21b437fe95dd9).rankingScript"
+rankprofile[].fef.property[].value "log10(rankingExpression(dvCountToUse) + 1)"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p@c05478688f81fe20).rankingScript"
+rankprofile[].fef.property[].value "log10(rankingExpression(uniqueRCountToUse) + 1)"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p@53f0a2c000e82f4).rankingScript"
+rankprofile[].fef.property[].value "log10(rankingExpression(uvCountToUse) + 1)"
+rankprofile[].fef.property[].name "rankingExpression(log10_1p@d7da61ad34902e89).rankingScript"
+rankprofile[].fef.property[].value "log10(rankingExpression(totalvote) + 1)"
+rankprofile[].fef.property[].name "rankingExpression(nn_input).rankingScript"
+rankprofile[].fef.property[].value "concat(rankingExpression(log10_1p@af9a8c53ba738798), concat(rankingExpression(log10_1p@6ad21b437fe95dd9), concat(rankingExpression(log10_1p@c05478688f81fe20), concat(rankingExpression(log10_1p@53f0a2c000e82f4), concat(rankingExpression(phat), concat(rankingExpression(log10_1p@d7da61ad34902e89), concat(rankingExpression(hsScoreToUse), concat(rankingExpression(tdToUse), rankingExpression(tScoreToUse), x), x), x), x), x), x), x), x)"
+rankprofile[].fef.property[].name "rankingExpression(nn_input).type"
+rankprofile[].fef.property[].value "tensor(x[9])"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights).rankingScript"
+rankprofile[].fef.property[].value "if (query(field) == 0, constant(field), query(field))"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@1f2b4afc2c45fbee).rankingScript"
+rankprofile[].fef.property[].value "if (query(W_0) == 0, constant(W_0), query(W_0))"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@e752cecc7900ff3e).rankingScript"
+rankprofile[].fef.property[].value "if (query(b_0) == 0, constant(b_0), query(b_0))"
+rankprofile[].fef.property[].name "rankingExpression(layer_0).rankingScript"
+rankprofile[].fef.property[].value "elu(join(reduce(join(rankingExpression(nn_input), rankingExpression(get_model_weights@1f2b4afc2c45fbee), f(a,b)(a * b)), sum, x), rankingExpression(get_model_weights@e752cecc7900ff3e), f(a,b)(a + b)))"
+rankprofile[].fef.property[].name "rankingExpression(layer_0).type"
+rankprofile[].fef.property[].value "tensor(hidden[9])"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@eac265fa16b752cf).rankingScript"
+rankprofile[].fef.property[].value "if (query(W_1) == 0, constant(W_1), query(W_1))"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@b953c19adb7d2154).rankingScript"
+rankprofile[].fef.property[].value "if (query(b_1) == 0, constant(b_1), query(b_1))"
+rankprofile[].fef.property[].name "rankingExpression(layer_1).rankingScript"
+rankprofile[].fef.property[].value "elu(join(reduce(join(rankingExpression(layer_0), rankingExpression(get_model_weights@eac265fa16b752cf), f(a,b)(a * b)), sum, hidden), rankingExpression(get_model_weights@b953c19adb7d2154), f(a,b)(a + b)))"
+rankprofile[].fef.property[].name "rankingExpression(layer_1).type"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@418462473aa32b7d).rankingScript"
+rankprofile[].fef.property[].value "if (query(W_out) == 0, constant(W_out), query(W_out))"
+rankprofile[].fef.property[].name "rankingExpression(get_model_weights@23f46853cab72961).rankingScript"
+rankprofile[].fef.property[].value "if (query(b_out) == 0, constant(b_out), query(b_out))"
+rankprofile[].fef.property[].name "rankingExpression(layer_out).rankingScript"
+rankprofile[].fef.property[].value "reduce(join(reduce(join(rankingExpression(layer_1), rankingExpression(get_model_weights@418462473aa32b7d), f(a,b)(a * b)), sum, out), rankingExpression(get_model_weights@23f46853cab72961), f(a,b)(a + b)), sum)"
+rankprofile[].fef.property[].name "vespa.rank.firstphase"
+rankprofile[].fef.property[].value "rankingExpression(freshnessRank)"
+rankprofile[].fef.property[].name "vespa.rank.secondphase"
+rankprofile[].fef.property[].value "rankingExpression(layer_out)"
+rankprofile[].fef.property[].name "vespa.hitcollector.heapsize"
+rankprofile[].fef.property[].value "2000"
+rankprofile[].fef.property[].name "vespa.type.query.b_out"
+rankprofile[].fef.property[].value "tensor(out[1])"
+rankprofile[].fef.property[].name "vespa.type.query.W_out"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_0"
+rankprofile[].fef.property[].value "tensor(hidden[9])"
+rankprofile[].fef.property[].name "vespa.type.query.b_1"
+rankprofile[].fef.property[].value "tensor(out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_1"
+rankprofile[].fef.property[].value "tensor(hidden[9],out[9])"
+rankprofile[].fef.property[].name "vespa.type.query.W_0"
+rankprofile[].fef.property[].value "tensor(hidden[9],x[9])"
diff --git a/config-model/src/test/integration/vespa/models/example.model b/config-model/src/test/integration/vespa/models/example.model
index 9579be4e44c..e9725d14923 100644
--- a/config-model/src/test/integration/vespa/models/example.model
+++ b/config-model/src/test/integration/vespa/models/example.model
@@ -19,7 +19,7 @@ model example {
     }
 
     function foo2() {
-        expression: max(sum(input1 * input2, name) * constant1asLarge, x) * constant2
+        expression: max(sum(input1 * input2, name) * constant(constant1asLarge), x) * constant2
     }
 
 }
\ No newline at end of file
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/RankingExpressionLoopDetectionTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/RankingExpressionLoopDetectionTestCase.java
index 0ff8a5cc7ca..9a0dcc7dd07 100644
--- a/config-model/src/test/java/com/yahoo/searchdefinition/RankingExpressionLoopDetectionTestCase.java
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/RankingExpressionLoopDetectionTestCase.java
@@ -40,7 +40,7 @@ public class RankingExpressionLoopDetectionTestCase {
             fail("Excepted exception");
         }
         catch (IllegalArgumentException e) {
-            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: Invocation loop: foo -> foo",
+            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: foo is invalid: Invocation loop: foo -> foo",
                          Exceptions.toMessageString(e));
         }
     }
@@ -75,7 +75,7 @@ public class RankingExpressionLoopDetectionTestCase {
             fail("Excepted exception");
         }
         catch (IllegalArgumentException e) {
-            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: Invocation loop: arg(5) -> foo -> arg(5)",
+            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: arg(5) is invalid: foo is invalid: arg(5) is invalid: Invocation loop: arg(5) -> foo -> arg(5)",
                          Exceptions.toMessageString(e));
         }
     }
@@ -110,7 +110,7 @@ public class RankingExpressionLoopDetectionTestCase {
             fail("Excepted exception");
         }
         catch (IllegalArgumentException e) {
-            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: Invocation loop: arg(foo) -> foo -> arg(foo)",
+            assertEquals("In search definition 'test', rank profile 'test': The function 'foo' is invalid: arg(foo) is invalid: a1 is invalid: foo is invalid: arg(foo) is invalid: Invocation loop: arg(foo) -> foo -> arg(foo)",
                          Exceptions.toMessageString(e));
         }
     }
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/derived/AbstractExportingTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/derived/AbstractExportingTestCase.java
index ef99ec28686..7fbca88cb61 100644
--- a/config-model/src/test/java/com/yahoo/searchdefinition/derived/AbstractExportingTestCase.java
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/derived/AbstractExportingTestCase.java
@@ -3,6 +3,7 @@ package com.yahoo.searchdefinition.derived;
 
 import com.yahoo.document.DocumenttypesConfig;
 import com.yahoo.document.config.DocumentmanagerConfig;
+import com.yahoo.search.query.profile.QueryProfileRegistry;
 import com.yahoo.searchdefinition.Search;
 import com.yahoo.searchdefinition.SearchBuilder;
 import com.yahoo.searchdefinition.SearchDefinitionTestCase;
@@ -29,11 +30,10 @@ public abstract class AbstractExportingTestCase extends SearchDefinitionTestCase
         deleteContent(toDir);
 
         SearchBuilder builder = SearchBuilder.createFromDirectory(searchDefRoot + dirName + "/");
-        //SearchBuilder builder = SearchBuilder.createFromFile(searchDefDir + name + ".sd");
         return derive(dirName, searchDefinitionName, builder);
     }
 
-    protected DerivedConfiguration derive(String dirName, String searchDefinitionName, SearchBuilder builder) throws IOException {
+    private DerivedConfiguration derive(String dirName, String searchDefinitionName, SearchBuilder builder) throws IOException {
         DerivedConfiguration config = new DerivedConfiguration(builder.getSearch(searchDefinitionName),
                                                                builder.getRankProfileRegistry(),
                                                                builder.getQueryProfileRegistry(),
@@ -85,14 +85,14 @@ public abstract class AbstractExportingTestCase extends SearchDefinitionTestCase
      * Asserts config is correctly derived given a builder.
      * This will fail if the builder contains multiple search definitions.
      */
-    protected DerivedConfiguration assertCorrectDeriving(SearchBuilder builder, String dirName) throws IOException, ParseException {
+    protected DerivedConfiguration assertCorrectDeriving(SearchBuilder builder, String dirName) throws IOException {
         builder.build();
         DerivedConfiguration derived = derive(dirName, null, builder);
         assertCorrectConfigFiles(dirName);
         return derived;
     }
 
-    protected DerivedConfiguration assertCorrectDeriving(SearchBuilder builder, Search search, String name) throws IOException, ParseException {
+    protected DerivedConfiguration assertCorrectDeriving(SearchBuilder builder, Search search, String name) throws IOException {
         DerivedConfiguration derived = derive(name, builder, search);
         assertCorrectConfigFiles(name);
         return derived;
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/derived/NeuralNetTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/derived/NeuralNetTestCase.java
new file mode 100644
index 00000000000..b299c7fa299
--- /dev/null
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/derived/NeuralNetTestCase.java
@@ -0,0 +1,16 @@
+// Copyright 2019 Oath Inc. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
+package com.yahoo.searchdefinition.derived;
+
+import com.yahoo.searchdefinition.parser.ParseException;
+import org.junit.Test;
+
+import java.io.IOException;
+
+public class NeuralNetTestCase extends AbstractExportingTestCase {
+
+    @Test
+    public void testNeuralNet() throws IOException, ParseException {
+        assertCorrectDeriving("neuralnet");
+    }
+
+}
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionTypeResolverTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionTypeResolverTestCase.java
index 1b917b6f3a3..3b3ce712387 100644
--- a/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionTypeResolverTestCase.java
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionTypeResolverTestCase.java
@@ -103,7 +103,9 @@ public class RankingExpressionTypeResolverTestCase {
             fail("Expected exception");
         }
         catch (IllegalArgumentException expected) {
-            assertEquals("In search definition 'test', rank profile 'my_rank_profile': The first-phase expression is invalid: An if expression must produce compatible types in both alternatives, but the 'true' type is tensor(x[],y[]) while the 'false' type is tensor(z[10])",
+            assertEquals("In search definition 'test', rank profile 'my_rank_profile': The first-phase expression is invalid: An if expression must produce compatible types in both alternatives, but the 'true' type is tensor(x[],y[]) while the 'false' type is tensor(z[10])" +
+                         "\n'true' branch: attribute(a)" +
+                         "\n'false' branch: attribute(b)",
                          Exceptions.toMessageString(expected));
         }
     }
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionsTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionsTestCase.java
index d4fcd09e249..1a7eb96483e 100644
--- a/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionsTestCase.java
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/processing/RankingExpressionsTestCase.java
@@ -23,8 +23,7 @@ public class RankingExpressionsTestCase extends SearchDefinitionTestCase {
     public void testFunctions() throws IOException, ParseException {
         RankProfileRegistry rankProfileRegistry = new RankProfileRegistry();
         Search search = SearchBuilder.createFromDirectory("src/test/examples/rankingexpressionfunction",
-                                                          rankProfileRegistry,
-                                                          new QueryProfileRegistry()).getSearch();
+                                                          rankProfileRegistry).getSearch();
         RankProfile functionsRankProfile = rankProfileRegistry.get(search, "macros");
         Map<String, RankProfile.RankingExpressionFunction> functions = functionsRankProfile.getFunctions();
         assertEquals(2, functions.get("titlematch$").function().arguments().size());
@@ -62,9 +61,7 @@ public class RankingExpressionsTestCase extends SearchDefinitionTestCase {
     @Test(expected = IllegalArgumentException.class)
     public void testThatIncludingFileInSubdirFails() throws IOException, ParseException {
         RankProfileRegistry registry = new RankProfileRegistry();
-        Search search = SearchBuilder.createFromDirectory("src/test/examples/rankingexpressioninfile",
-                                                          registry,
-                                                          new QueryProfileRegistry()).getSearch();
+        Search search = SearchBuilder.createFromDirectory("src/test/examples/rankingexpressioninfile", registry).getSearch();
         new DerivedConfiguration(search, registry, new QueryProfileRegistry(), new ImportedMlModels()); // rank profile parsing happens during deriving
     }
 
diff --git a/config-model/src/test/java/com/yahoo/searchdefinition/processing/TensorTransformTestCase.java b/config-model/src/test/java/com/yahoo/searchdefinition/processing/TensorTransformTestCase.java
index fe150b51961..15c1d24ce33 100644
--- a/config-model/src/test/java/com/yahoo/searchdefinition/processing/TensorTransformTestCase.java
+++ b/config-model/src/test/java/com/yahoo/searchdefinition/processing/TensorTransformTestCase.java
@@ -58,8 +58,8 @@ public class TensorTransformTestCase extends SearchDefinitionTestCase {
                                     "max(attribute(tensor_field_1),x)");
         assertTransformedExpression("1+reduce(attribute(tensor_field_1),max,x)",
                                     "1 + max(attribute(tensor_field_1),x)");
-        assertTransformedExpression("if(attribute(double_field),1+reduce(attribute(tensor_field_1),max,x),attribute(tensor_field_1))",
-                                    "if(attribute(double_field),1 + max(attribute(tensor_field_1),x),attribute(tensor_field_1))");
+        assertTransformedExpression("if(attribute(double_field),1+reduce(attribute(tensor_field_1),max,x),reduce(attribute(tensor_field_1),sum,x))",
+                                    "if(attribute(double_field),1 + max(attribute(tensor_field_1),x),reduce(attribute(tensor_field_1), sum, x))");
         assertTransformedExpression("reduce(max(attribute(tensor_field_1),attribute(tensor_field_2)),max,x)",
                                     "max(max(attribute(tensor_field_1),attribute(tensor_field_2)),x)");
         assertTransformedExpression("reduce(if(attribute(double_field),attribute(tensor_field_2),attribute(tensor_field_2)),max,x)",
diff --git a/container-search/src/main/java/com/yahoo/search/query/profile/config/QueryProfileXMLReader.java b/container-search/src/main/java/com/yahoo/search/query/profile/config/QueryProfileXMLReader.java
index eb4a0ad6be4..210b4899c58 100644
--- a/container-search/src/main/java/com/yahoo/search/query/profile/config/QueryProfileXMLReader.java
+++ b/container-search/src/main/java/com/yahoo/search/query/profile/config/QueryProfileXMLReader.java
@@ -23,7 +23,7 @@ import java.util.Collections;
 import java.util.List;
 
 /**
- * A class which imports query profiles and types from XML files
+ * Importer of query profiles and types from XML files
  *
  * @author bratseth
  */
diff --git a/documentapi/src/main/java/com/yahoo/documentapi/VisitorParameters.java b/documentapi/src/main/java/com/yahoo/documentapi/VisitorParameters.java
index 82e7a87e95b..df0e0f0abdd 100644
--- a/documentapi/src/main/java/com/yahoo/documentapi/VisitorParameters.java
+++ b/documentapi/src/main/java/com/yahoo/documentapi/VisitorParameters.java
@@ -36,7 +36,7 @@ public class VisitorParameters extends Parameters {
     private String remoteDataHandler = null;
     private VisitorDataHandler localDataHandler;
     private VisitorControlHandler controlHandler;
-    private Map<String, byte []> libraryParameters = new TreeMap<String, byte []>();
+    private Map<String, byte []> libraryParameters = new TreeMap<>();
     private Route visitRoute = null;
     private float weight = 1;
     private long maxFirstPassHits = -1;
diff --git a/model-integration/src/test/java/ai/vespa/rankingexpression/importer/vespa/VespaImportTestCase.java b/model-integration/src/test/java/ai/vespa/rankingexpression/importer/vespa/VespaImportTestCase.java
index 25a24792432..c7210e6710a 100644
--- a/model-integration/src/test/java/ai/vespa/rankingexpression/importer/vespa/VespaImportTestCase.java
+++ b/model-integration/src/test/java/ai/vespa/rankingexpression/importer/vespa/VespaImportTestCase.java
@@ -40,7 +40,7 @@ public class VespaImportTestCase {
         assertEquals(2, model.expressions().size());
         assertEquals("reduce(reduce(input1 * input2, sum, name) * constant1, max, x) * constant2",
                      model.expressions().get("foo1").getRoot().toString());
-        assertEquals("reduce(reduce(input1 * input2, sum, name) * constant1asLarge, max, x) * constant2",
+        assertEquals("reduce(reduce(input1 * input2, sum, name) * constant(constant1asLarge), max, x) * constant2",
                      model.expressions().get("foo2").getRoot().toString());
 
         List<ImportedMlFunction> functions = model.outputExpressions();
diff --git a/model-integration/src/test/models/vespa/example.model b/model-integration/src/test/models/vespa/example.model
index 6d660732db9..269ed83b695 100644
--- a/model-integration/src/test/models/vespa/example.model
+++ b/model-integration/src/test/models/vespa/example.model
@@ -19,7 +19,7 @@ model example {
     }
 
     function foo2() {
-        expression: reduce(sum(input1 * input2, name) * constant1asLarge, max, x) * constant2
+        expression: reduce(sum(input1 * input2, name) * constant(constant1asLarge), max, x) * constant2
     }
 
 }
\ No newline at end of file
diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/FunctionNode.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/FunctionNode.java
index c4f3a75f2f8..2aedec2109b 100755
--- a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/FunctionNode.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/FunctionNode.java
@@ -5,6 +5,7 @@ import com.yahoo.searchlib.rankingexpression.Reference;
 import com.yahoo.searchlib.rankingexpression.evaluation.Context;
 import com.yahoo.searchlib.rankingexpression.evaluation.DoubleValue;
 import com.yahoo.searchlib.rankingexpression.evaluation.Value;
+import com.yahoo.searchlib.rankingexpression.transform.TensorMaxMinTransformer;
 import com.yahoo.tensor.TensorType;
 import com.yahoo.tensor.evaluation.TypeContext;
 import com.yahoo.tensor.functions.Join;
@@ -67,6 +68,11 @@ public final class FunctionNode extends CompositeNode {
 
     @Override
     public TensorType type(TypeContext<Reference> context) {
+        // Check if this node should be interpreted as tensor reduce, as this impacts the type
+        ExpressionNode thisTransformed = TensorMaxMinTransformer.transformFunctionNode(this, context);
+        if (thisTransformed != this)
+            return thisTransformed.type(context);
+
         if (arguments.expressions().size() == 0)
             return TensorType.empty;
 
diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/IfNode.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/IfNode.java
index 28dc623be72..92c6d6f8638 100755
--- a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/IfNode.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/IfNode.java
@@ -85,7 +85,9 @@ public final class IfNode extends CompositeNode {
         return trueType.dimensionwiseGeneralizationWith(falseType).orElseThrow(() ->
             new IllegalArgumentException("An if expression must produce compatible types in both " +
                                          "alternatives, but the 'true' type is " + trueType + " while the " +
-                                         "'false' type is " + falseType)
+                                         "'false' type is " + falseType +
+                                         "\n'true' branch: " + trueExpression +
+                                         "\n'false' branch: " + falseExpression)
         );
     }
 
diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/ReferenceNode.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/ReferenceNode.java
index eb8d2229a6d..e15ce158e83 100755
--- a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/ReferenceNode.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/rule/ReferenceNode.java
@@ -95,7 +95,13 @@ public final class ReferenceNode extends CompositeNode {
 
     @Override
     public TensorType type(TypeContext<Reference> context) {
-        TensorType type = context.getType(reference);
+        TensorType type = null;
+        try {
+            type = context.getType(reference);
+        }
+        catch (IllegalArgumentException e) {
+            throw new IllegalArgumentException(reference + " is invalid", e);
+        }
         if (type == null)
             throw new IllegalArgumentException("Unknown feature '" + toString() + "'");
         return type;
diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/ExpressionTransformer.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/ExpressionTransformer.java
index 22d314bcb28..31567ba120b 100644
--- a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/ExpressionTransformer.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/ExpressionTransformer.java
@@ -10,7 +10,7 @@ import java.util.List;
 
 /**
  * Superclass of expression transformers. The scope (lifetime) of a transformer instance is a single compilation
- * of alle the expressions in one rank profile.
+ * of all the expressions in one rank profile.
  *
  * @author bratseth
  */
diff --git a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/TensorTransformer.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TensorMaxMinTransformer.java
index 5d03c323803..979c5b0f88c 100644
--- a/config-model/src/main/java/com/yahoo/searchdefinition/expressiontransforms/TensorTransformer.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TensorMaxMinTransformer.java
@@ -1,54 +1,40 @@
 // Copyright 2017 Yahoo Holdings. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
-package com.yahoo.searchdefinition.expressiontransforms;
+package com.yahoo.searchlib.rankingexpression.transform;
 
-import com.yahoo.searchdefinition.RankProfile;
-import com.yahoo.searchdefinition.RankingConstant;
-import com.yahoo.searchdefinition.document.Attribute;
 import com.yahoo.searchlib.rankingexpression.Reference;
-import com.yahoo.searchlib.rankingexpression.evaluation.Context;
-import com.yahoo.searchlib.rankingexpression.evaluation.DoubleValue;
-import com.yahoo.searchlib.rankingexpression.evaluation.MapContext;
-import com.yahoo.searchlib.rankingexpression.evaluation.StringValue;
-import com.yahoo.searchlib.rankingexpression.evaluation.TensorValue;
-import com.yahoo.searchlib.rankingexpression.evaluation.Value;
 import com.yahoo.searchlib.rankingexpression.rule.CompositeNode;
 import com.yahoo.searchlib.rankingexpression.rule.ExpressionNode;
 import com.yahoo.searchlib.rankingexpression.rule.FunctionNode;
 import com.yahoo.searchlib.rankingexpression.rule.NameNode;
 import com.yahoo.searchlib.rankingexpression.rule.ReferenceNode;
 import com.yahoo.searchlib.rankingexpression.rule.TensorFunctionNode;
-import com.yahoo.searchlib.rankingexpression.transform.ExpressionTransformer;
-import com.yahoo.tensor.Tensor;
 import com.yahoo.tensor.TensorType;
 import com.yahoo.tensor.evaluation.TypeContext;
 import com.yahoo.tensor.functions.Reduce;
 
-import java.util.List;
 import java.util.Optional;
 
 /**
- * Transforms and simplifies tensor expressions.
- *
- * Currently transforms min(tensor,dim) and max(tensor,dim) to
+ * Transforms min(tensor,dim) and max(tensor,dim) to
  * reduce(tensor,min/max,dim). This is necessary as the backend does
  * not recognize these forms of min and max.
  *
  * @author lesters
  */
-public class TensorTransformer extends ExpressionTransformer<RankProfileTransformContext> {
+public class TensorMaxMinTransformer<CONTEXT extends TransformContext> extends ExpressionTransformer<CONTEXT> {
 
     @Override
-    public ExpressionNode transform(ExpressionNode node, RankProfileTransformContext context) {
+    public ExpressionNode transform(ExpressionNode node, CONTEXT context) {
         if (node instanceof CompositeNode) {
             node = transformChildren((CompositeNode) node, context);
         }
         if (node instanceof FunctionNode) {
-            node = transformFunctionNode((FunctionNode) node, context);
+            node = transformFunctionNode((FunctionNode) node, context.types());
         }
         return node;
     }
 
-    private ExpressionNode transformFunctionNode(FunctionNode node, RankProfileTransformContext context) {
+    public static ExpressionNode transformFunctionNode(FunctionNode node, TypeContext<Reference> context) {
         switch (node.getFunction()) {
             case min:
             case max:
@@ -62,14 +48,14 @@ public class TensorTransformer extends ExpressionTransformer<RankProfileTransfor
      * argument returns a tensor type and the second argument is a valid
      * dimension in the tensor.
      */
-    private ExpressionNode transformMaxAndMinFunctionNode(FunctionNode node, RankProfileTransformContext context) {
+    private static ExpressionNode transformMaxAndMinFunctionNode(FunctionNode node, TypeContext<Reference> context) {
         if (node.children().size() != 2) {
             return node;
         }
         ExpressionNode arg1 = node.children().get(0);
         Optional<String> dimension = dimensionName(node.children().get(1));
         if (dimension.isPresent()) {
-            TensorType type = arg1.type(context.types());
+            TensorType type = arg1.type(context);
             if (type.dimension(dimension.get()).isPresent()) {
                 return replaceMaxAndMinFunction(node);
             }
@@ -77,7 +63,7 @@ public class TensorTransformer extends ExpressionTransformer<RankProfileTransfor
         return node;
     }
 
-    private Optional<String> dimensionName(ExpressionNode node) {
+    private static Optional<String> dimensionName(ExpressionNode node) {
         if (node instanceof ReferenceNode) {
             Reference reference = ((ReferenceNode)node).reference();
             if (reference.isIdentifier())
@@ -93,7 +79,7 @@ public class TensorTransformer extends ExpressionTransformer<RankProfileTransfor
         }
     }
 
-    private ExpressionNode replaceMaxAndMinFunction(FunctionNode node) {
+    private static ExpressionNode replaceMaxAndMinFunction(FunctionNode node) {
         ExpressionNode arg1 = node.children().get(0);
         ExpressionNode arg2 = node.children().get(1);
 
diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TransformContext.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TransformContext.java
index 7485ce69f98..0113a650277 100644
--- a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TransformContext.java
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TransformContext.java
@@ -1,7 +1,9 @@
 // Copyright 2018 Yahoo Holdings. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
 package com.yahoo.searchlib.rankingexpression.transform;
 
+import com.yahoo.searchlib.rankingexpression.Reference;
 import com.yahoo.searchlib.rankingexpression.evaluation.Value;
+import com.yahoo.tensor.evaluation.TypeContext;
 
 import java.util.Map;
 
@@ -13,11 +15,19 @@ import java.util.Map;
 public class TransformContext {
 
     private final Map<String, Value> constants;
+    private final TypeContext<Reference> types;
 
-    public TransformContext(Map<String, Value> constants) {
+    public TransformContext(Map<String, Value> constants, TypeContext<Reference> types) {
         this.constants = constants;
+        this.types = types;
     }
 
     public Map<String, Value> constants() { return constants; }
 
+    /**
+     * Returns the types known in this context. We may have type information for references
+     * for which no value is available
+     */
+    public TypeContext<Reference> types() { return types; }
+
 }
diff --git a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/evaluation/TypeResolutionTestCase.java b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/evaluation/TypeResolutionTestCase.java
index a08d510eec4..88838b5aed0 100644
--- a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/evaluation/TypeResolutionTestCase.java
+++ b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/evaluation/TypeResolutionTestCase.java
@@ -53,7 +53,9 @@ public class TypeResolutionTestCase {
         }
         catch (IllegalArgumentException expected) {
             assertEquals("An if expression must produce compatible types in both alternatives, " +
-                         "but the 'true' type is tensor(x[]) while the 'false' type is tensor(y[])",
+                         "but the 'true' type is tensor(x[]) while the 'false' type is tensor(y[])" +
+                         "\n'true' branch: query(x1)" +
+                         "\n'false' branch: query(y1)",
                          expected.getMessage());
         }
         catch (ParseException e) {
diff --git a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/ConstantDereferencerTestCase.java b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/ConstantDereferencerTestCase.java
index 1f28f0b0129..a41fb02f784 100644
--- a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/ConstantDereferencerTestCase.java
+++ b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/ConstantDereferencerTestCase.java
@@ -2,6 +2,7 @@
 package com.yahoo.searchlib.rankingexpression.transform;
 
 import com.yahoo.searchlib.rankingexpression.RankingExpression;
+import com.yahoo.searchlib.rankingexpression.evaluation.MapTypeContext;
 import com.yahoo.searchlib.rankingexpression.evaluation.Value;
 import com.yahoo.searchlib.rankingexpression.parser.ParseException;
 import org.junit.Test;
@@ -24,7 +25,7 @@ public class ConstantDereferencerTestCase {
         constants.put("a", Value.parse("1.0"));
         constants.put("b", Value.parse("2"));
         constants.put("c", Value.parse("3.5"));
-        TransformContext context = new TransformContext(constants);
+        TransformContext context = new TransformContext(constants, new MapTypeContext());
 
         assertEquals("1.0 + 2.0 + 3.5", c.transform(new RankingExpression("a + b + c"), context).toString());
         assertEquals("myFunction(1.0,2.0)", c.transform(new RankingExpression("myFunction(a, b)"), context).toString());
diff --git a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/SimplifierTestCase.java b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/SimplifierTestCase.java
index 8fac3395ac0..f4b1b0ceee2 100644
--- a/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/SimplifierTestCase.java
+++ b/searchlib/src/test/java/com/yahoo/searchlib/rankingexpression/transform/SimplifierTestCase.java
@@ -1,9 +1,11 @@
 // Copyright 2017 Yahoo Holdings. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
 package com.yahoo.searchlib.rankingexpression.transform;
 
+import com.yahoo.log.event.Collection;
 import com.yahoo.searchlib.rankingexpression.RankingExpression;
 import com.yahoo.searchlib.rankingexpression.evaluation.Context;
 import com.yahoo.searchlib.rankingexpression.evaluation.MapContext;
+import com.yahoo.searchlib.rankingexpression.evaluation.MapTypeContext;
 import com.yahoo.searchlib.rankingexpression.parser.ParseException;
 import com.yahoo.searchlib.rankingexpression.rule.CompositeNode;
 import org.junit.Test;
@@ -20,7 +22,7 @@ public class SimplifierTestCase {
     @Test
     public void testSimplify() throws ParseException {
         Simplifier s = new Simplifier();
-        TransformContext c = new TransformContext(Collections.emptyMap());
+        TransformContext c = new TransformContext(Collections.emptyMap(), new MapTypeContext());
         assertEquals("a + b", s.transform(new RankingExpression("a + b"), c).toString());
         assertEquals("6.5", s.transform(new RankingExpression("1.0 + 2.0 + 3.5"), c).toString());
         assertEquals("6.5", s.transform(new RankingExpression("1.0 + ( 2.0 + 3.5 )"), c).toString());
@@ -45,7 +47,7 @@ public class SimplifierTestCase {
     @Test
     public void testSimplifyComplexExpression() throws ParseException {
         RankingExpression initial = new RankingExpression("sqrt(if (if (INFERRED * 0.9 < INFERRED, GMP, (1 + 1.1) * INFERRED) < INFERRED * INFERRED - INFERRED, if (GMP < 85.80799542793133 * GMP, INFERRED, if (GMP < GMP, tanh(INFERRED), log(76.89956221113943))), tanh(tanh(INFERRED))) * sqrt(sqrt(GMP + INFERRED)) * GMP ) + 13.5 * (1 - GMP) * pow(GMP * 0.1, 2 + 1.1 * 0)");
-        TransformContext c = new TransformContext(Collections.emptyMap());
+        TransformContext c = new TransformContext(Collections.emptyMap(), new MapTypeContext());
         RankingExpression simplified = new Simplifier().transform(initial, c);
 
         Context context = new MapContext();
@@ -70,7 +72,7 @@ public class SimplifierTestCase {
     @Test
     public void testParenthesisPreservation() throws ParseException {
         Simplifier s = new Simplifier();
-        TransformContext c = new TransformContext(Collections.emptyMap());
+        TransformContext c = new TransformContext(Collections.emptyMap(), new MapTypeContext());
         CompositeNode transformed = (CompositeNode)s.transform(new RankingExpression("a + (b + c) / 100000000.0"), c).getRoot();
         assertEquals("a + (b + c) / 100000000.0", transformed.toString());
     }