Merge pull request #139 from tigergraph/3.9

Updating master to 3.9
tigergraph · Mar 7, 2023 · 4d91fb1 · 4d91fb1
2 parents 8383057 + 9d4a391
commit 4d91fb1
Show file tree

Hide file tree

Showing 9 changed files with 84 additions and 40 deletions.
diff --git a/.github/auto_request_review.yml b/.github/auto_request_review.yml
@@ -11,16 +11,17 @@ reviewers:
     group1:
       - yimingpantg
       - TannerW
-      - wyatt-joyner-tg
-      - parkererickson
-      - lennessyy
-      - a-m-thomas
+      - wyattjoynertg
+      - parkererickson-tg
+      - alexthomasTG
     secondary:
       - yimingpantg
-      - wyatt-joyner-tg
+      - wyattjoynertg
+      - TannerW
     tertiary:
       - xinyuchtg
-      - victor-gsl
+      - victorleeTG
+      - TannerW
 
 files:
   # All review request will be sent to the following groups/people.

diff --git a/.github/workflows/review.yml b/.github/workflows/review.yml
@@ -12,6 +12,7 @@ jobs:
   auto-request-review:
     runs-on: ubuntu-latest
     steps:
-      - uses: necojackarc/auto-request-review@v0.4.0
+      - uses: necojackarc/auto-request-review@v0.10.0
         with:
           token: ${{ secrets.GITHUB_TOKEN }}
+          config: .github/auto_request_review.yml
diff --git a/GDBMS_ALGO/centrality/pagerank.gsql b/GDBMS_ALGO/centrality/pagerank.gsql
@@ -56,6 +56,7 @@ CREATE TEMPLATE QUERY GDBMS_ALGO.centrality.pagerank (STRING v_type, STRING e_ty
 
 TYPEDEF TUPLE<VERTEX Vertex_ID, FLOAT score> Vertex_Score;
 HeapAccum<Vertex_Score>(top_k, score DESC) @@top_scores_heap;
+SetAccum<VERTEX> @@top_vertices;      # vertices with top score
 MaxAccum<FLOAT> @@max_diff = 9999;    # max score change in an iteration
 SumAccum<FLOAT> @sum_recvd_score = 0; # sum of scores each vertex receives FROM neighbors
 SumAccum<FLOAT> @sum_score = 1;           # initial score for every vertex is 1.
@@ -99,11 +100,19 @@ V = SELECT s
 IF print_results THEN
     PRINT @@top_scores_heap;
     IF display_edges THEN
-        PRINT Start[Start.@sum_score];
-	Start = SELECT s
-	        FROM Start:s -(e_type:e)- v_type:t
+
+        FOREACH vert IN @@top_scores_heap DO
+            @@top_vertices += vert.Vertex_ID;
+        END;
+
+        Top = {@@top_vertices};
+        Top = SELECT s
+	        FROM Top:s -(e_type:e)- v_type:t
+            WHERE @@top_vertices.contains(t)
 	        ACCUM @@edge_set += e;
+
         PRINT @@edge_set;
+        PRINT Top;
     END;
 END;
 }
diff --git a/GDBMS_ALGO/centrality/pagerank_wt.gsql b/GDBMS_ALGO/centrality/pagerank_wt.gsql
@@ -60,6 +60,7 @@ CREATE TEMPLATE QUERY GDBMS_ALGO.centrality.pagerank_wt (STRING v_type, STRING e
 
 TYPEDEF TUPLE<VERTEX Vertex_ID, FLOAT score> Vertex_Score;
 HeapAccum<Vertex_Score>(top_k, score DESC) @@top_scores_heap;
+SetAccum<VERTEX> @@top_vertices;      # vertices with top score
 MaxAccum<FLOAT> @@max_diff = 9999;    # max score change in an iteration
 SumAccum<FLOAT> @sum_recvd_score = 0; # sum of scores each vertex receives FROM neighbors
 SumAccum<FLOAT> @sum_score = 1;           # initial score for every vertex is 1.
@@ -107,11 +108,19 @@ V = SELECT s
 IF print_results THEN
     PRINT @@top_scores_heap;
     IF display_edges THEN
-        PRINT Start[Start.@sum_score];
-	Start = SELECT s
-		FROM Start:s -(e_type:e)- v_type:t
-		ACCUM @@edge_set += e;
-	PRINT @@edge_set;
+
+        FOREACH vert IN @@top_scores_heap DO
+            @@top_vertices += vert.Vertex_ID;
+        END;
+
+        Top = {@@top_vertices};
+        Top = SELECT s
+	        FROM Top:s -(e_type:e)- v_type:t
+            WHERE @@top_vertices.contains(t)
+	        ACCUM @@edge_set += e;
+
+        PRINT @@edge_set;
+        PRINT Top;
     END;
 END;
 }
diff --git a/algorithms/Centrality/pagerank/global/unweighted/tg_pagerank.gsql b/algorithms/Centrality/pagerank/global/unweighted/tg_pagerank.gsql
@@ -56,6 +56,7 @@ CREATE QUERY tg_pagerank (STRING v_type, STRING e_type,
 
 TYPEDEF TUPLE<VERTEX Vertex_ID, FLOAT score> Vertex_Score;
 HeapAccum<Vertex_Score>(top_k, score DESC) @@top_scores_heap;
+SetAccum<VERTEX> @@top_vertices;      # vertices with top score
 MaxAccum<FLOAT> @@max_diff = 9999;    # max score change in an iteration
 SumAccum<FLOAT> @sum_recvd_score = 0; # sum of scores each vertex receives FROM neighbors
 SumAccum<FLOAT> @sum_score = 1;           # initial score for every vertex is 1.
@@ -99,11 +100,19 @@ V = SELECT s
 IF print_results THEN
     PRINT @@top_scores_heap;
     IF display_edges THEN
-        PRINT Start[Start.@sum_score];
-	Start = SELECT s
-	        FROM Start:s -(e_type:e)- v_type:t
+
+        FOREACH vert IN @@top_scores_heap DO
+            @@top_vertices += vert.Vertex_ID;
+        END;
+
+        Top = {@@top_vertices};
+        Top = SELECT s
+	        FROM Top:s -(e_type:e)- v_type:t
+            WHERE @@top_vertices.contains(t)
 	        ACCUM @@edge_set += e;
+
         PRINT @@edge_set;
+        PRINT Top;
     END;
 END;
 }
diff --git a/algorithms/Centrality/pagerank/global/weighted/tg_pagerank_wt.gsql b/algorithms/Centrality/pagerank/global/weighted/tg_pagerank_wt.gsql
@@ -60,6 +60,7 @@ CREATE QUERY tg_pagerank_wt (STRING v_type, STRING e_type, STRING weight_attribu
 
 TYPEDEF TUPLE<VERTEX Vertex_ID, FLOAT score> Vertex_Score;
 HeapAccum<Vertex_Score>(top_k, score DESC) @@top_scores_heap;
+SetAccum<VERTEX> @@top_vertices;      # vertices with top score
 MaxAccum<FLOAT> @@max_diff = 9999;    # max score change in an iteration
 SumAccum<FLOAT> @sum_recvd_score = 0; # sum of scores each vertex receives FROM neighbors
 SumAccum<FLOAT> @sum_score = 1;           # initial score for every vertex is 1.
@@ -107,11 +108,19 @@ V = SELECT s
 IF print_results THEN
     PRINT @@top_scores_heap;
     IF display_edges THEN
-        PRINT Start[Start.@sum_score];
-	Start = SELECT s
-		FROM Start:s -(e_type:e)- v_type:t
-		ACCUM @@edge_set += e;
-	PRINT @@edge_set;
+
+        FOREACH vert IN @@top_scores_heap DO
+            @@top_vertices += vert.Vertex_ID;
+        END;
+
+        Top = {@@top_vertices};
+        Top = SELECT s
+	        FROM Top:s -(e_type:e)- v_type:t
+            WHERE @@top_vertices.contains(t)
+	        ACCUM @@edge_set += e;
+
+        PRINT @@edge_set;
+        PRINT Top;
     END;
 END;
 }
diff --git a/algorithms/GraphML/Embeddings/FastRP/tg_algo_fastRP.yml b/algorithms/GraphML/Embeddings/FastRP/tg_algo_fastRP.yml
@@ -12,8 +12,8 @@
     sha_id: ed6ea869749977cc0f3df71225d7325fb81c9767
     description: "Generates node embeddings of low dimensionality through random projections from the graph's adjacency matrix (a high-dimensional matrix) to a low-dimensional matrix, significantly reducing the computing power required for data processing. "
     schema_constraints: This algorithm can optionally write the embedding results back to the graph if the target vertex type(s) have a list attribute of type DOUBLE.
-    version: lib3.0
-    include: false
+    version: lib3.8
+    include: true
     dependencies:
       ExprFunctions:
         sha_id: ed6ea869749977cc0f3df71225d7325fb81c9767
diff --git a/algorithms/GraphML/Embeddings/FastRP/tg_fastRP.gsql b/algorithms/GraphML/Embeddings/FastRP/tg_fastRP.gsql
@@ -1,7 +1,7 @@
 CREATE QUERY tg_fastRP(
-  SET<STRING> v_type,
-  SET<STRING> e_type,
-  SET<STRING> output_v_type,
+  SET<STRING> v_type_set,
+  SET<STRING> e_type_set,
+  SET<STRING> output_v_type_set,
   STRING iteration_weights,
   FLOAT beta,
   INT embedding_dimension,
@@ -57,11 +57,11 @@ CREATE QUERY tg_fastRP(
 
 
   Parameters:
-    v_type:
+    v_type_set:
       Set of all vertex types to traverse in the graph.
-    e_type:
+    e_type_set:
       Set of all edge types to traverse in the graph.
-    output_v_type:
+    output_v_type_set:
       Set of vertex types to produce output embeddings for.
     iteration_weights:
       A comma-separated string of numbers to weight each iteration of the embedding process.
@@ -165,7 +165,7 @@ CREATE QUERY tg_fastRP(
     weight_idx = weight_idx + 1;
   END;
 
-  verts = {v_type};
+  verts = {v_type_set};
   // component_attr should indicate which batch each vertex belongs to, if none defined, just use all vertex types provided
   IF component_attribute != "" THEN
     verts =
@@ -182,22 +182,22 @@ CREATE QUERY tg_fastRP(
   // L stores the normalized diagonal elements of an inverse degree matrix
   // this is defined in the original algorithm and is useful for initialization
   verts =
-    SELECT s FROM verts:s -(e_type:e)- v_type:t
+    SELECT s FROM verts:s -(e_type_set:e)- v_type_set:t
     WHERE t.@include == TRUE
     ACCUM @@m += 1
-    POST-ACCUM s.@L = pow(s.outdegree(e_type) / @@m, beta);
+    POST-ACCUM s.@L = pow(s.outdegree(e_type_set) / @@m, beta);
 
   // randomly initialize the topological embedding regions
   verts =
-    SELECT s FROM verts:s -(e_type:e)- v_type:t
+    SELECT s FROM verts:s -(e_type_set:e)- v_type_set:t
     WHERE t.@include == TRUE
     ACCUM
       // PRNG code
       INT inc = (getvid(s)+_inc),
       INT r = ((inc+_mult*random_seed) % _mod),
       FLOAT mr = 0,
       STRING temp_e_type = e.type,
-      // if the edge type wasn't explicitly specified, but was provided in e_type,
+      // if the edge type wasn't explicitly specified, but was provided in e_type_set,
       // then its information will reside in the 'default' region of the embedding vector
       IF @@embedding_dim_map.containsKey(e.type) == FALSE THEN
         temp_e_type = "default"
@@ -221,14 +221,14 @@ CREATE QUERY tg_fastRP(
 
     // propagate embeddings to neighbors and normalize
     verts =
-      SELECT s FROM verts:s -(e_type)- v_type:t
+      SELECT s FROM verts:s -(e_type_set)- v_type_set:t
       WHERE t.@include == TRUE
       ACCUM
         t.@embedding_arr += s.@embedding_arr
       POST-ACCUM
         // first calculate square sum to help normalize
         FLOAT square_sum = 0,
-        FLOAT out = max([1.0, t.outdegree(e_type)]),
+        FLOAT out = max([1.0, t.outdegree(e_type_set)]),
         FOREACH (i,total) IN t.@embedding_arr DO
           square_sum = square_sum + pow(total / out, 2)
         END,
@@ -256,7 +256,7 @@ CREATE QUERY tg_fastRP(
   // the arrays are converted to lists for compatibility with GSQL's LIST type
   verts =
     SELECT s FROM verts:s
-    WHERE output_v_type.size() == 0 OR output_v_type.contains(s.type)
+    WHERE output_v_type_set.size() == 0 OR output_v_type_set.contains(s.type)
     POST-ACCUM
     FOREACH i IN RANGE[0, embedding_dimension-1] DO
       s.@final_embedding_list += s.@final_embedding_arr.get(i) / @@weights.size() // Average by # of iterations

diff --git a/manifest.json b/manifest.json
@@ -108,7 +108,13 @@
              "path": "algorithms/Community/connected_components/strongly_connected_components/standard/tg_scc.gsql",
              "schema_type": "VERTEX",
              "value_type": "INT"
-          }
+          },
+          "weakly_connected_components": {
+            "name": "tg_wcc",
+            "path": "algorithms/Community/connected_components/weakly_connected_components/standard/tg_wcc.gsql",
+            "schema_type": "VERTEX",
+            "value_type": "INT"
+         }
        },
        "k_core": {
           "name": "tg_kcore",