Add gmv aggregate

Signed-off-by: Andrew Stein <steinlink@gmail.com>

Author: texodus

rust/perspective-python/perspective/tests/table/test_view.py

@@ -424,6 +424,118 @@ def test_view_aggregate_mean(self):
             {"__ROW_PATH__": ["a"], "y": 300 / 2},
         ]
 
+    def test_view_aggregate_gmv(self):
+        data = {
+            "division": [
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+                "D1",
+                "D2",
+            ],
+            "trading area": [
+                "A",
+                "B",
+                "C",
+                "D",
+                "E",
+                "A",
+                "B",
+                "C",
+                "D",
+                "E",
+                "A",
+                "B",
+                "C",
+                "D",
+                "E",
+                "A",
+                "B",
+                "C",
+                "D",
+                "E",
+            ],
+            "symbol": [
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+                "MSFT",
+                "AAPL",
+                "GOOG",
+            ],
+            "MV": [
+                1500,
+                1200,
+                1300,
+                1400,
+                1600,
+                1100,
+                1700,
+                1800,
+                1900,
+                2000,
+                -2100,
+                -2200,
+                -2300,
+                -2400,
+                -2500,
+                -2600,
+                -2700,
+                -2800,
+                -2900,
+                -3000,
+            ],
+        }
+
+        tbl = Table(data)
+        view = tbl.view(
+            aggregates={"MV": "gmv"}, group_by=["division", "symbol"], columns=["MV"]
+        )
+
+        assert view.to_columns() == {
+            "__ROW_PATH__": [
+                [],
+                ["D1"],
+                ["D1", "AAPL"],
+                ["D1", "GOOG"],
+                ["D1", "MSFT"],
+                ["D2"],
+                ["D2", "AAPL"],
+                ["D2", "GOOG"],
+                ["D2", "MSFT"],
+            ],
+            "MV": [10000, 5900, -2000, -3200, 700, 7100, 800, -2400, -3900],
+        }
+
     def test_view_aggregate_mean_from_schema(self):
         data = [
             {"a": "a", "x": 1, "y": 200},

rust/perspective-server/cpp/perspective/src/cpp/aggspec.cpp

@@ -247,6 +247,9 @@ t_aggspec::agg_str() const {
         case AGGTYPE_STANDARD_DEVIATION: {
             return "stddev";
         }
+        case AGGTYPE_GMV: {
+            return "gmv";
+        }
         default: {
             PSP_COMPLAIN_AND_ABORT("Unknown agg type");
             return "unknown";
@@ -343,6 +346,7 @@ t_aggspec::get_output_specs(const t_schema& schema) const {
         case AGGTYPE_SUM:
         case AGGTYPE_SUM_ABS:
         case AGGTYPE_ABS_SUM:
+        case AGGTYPE_GMV:
         case AGGTYPE_PCT_SUM_PARENT:
         case AGGTYPE_PCT_SUM_GRAND_TOTAL:
         case AGGTYPE_MUL:

rust/perspective-server/cpp/perspective/src/cpp/base.cpp

@@ -589,6 +589,9 @@ str_to_aggtype(const std::string& str) {
     if (str == "stddev" || str == "standard deviation") {
         return t_aggtype::AGGTYPE_STANDARD_DEVIATION;
     }
+    if (str == "gmv") {
+        return t_aggtype::AGGTYPE_GMV;
+    }
 
     std::stringstream ss;
     ss << "Encountered unknown aggregate operation: '" << str << "'"

rust/perspective-server/cpp/perspective/src/cpp/config.cpp

@@ -226,6 +226,7 @@ t_config::setup(
             case AGGTYPE_SUM_NOT_NULL:
             case AGGTYPE_SUM_ABS:
             case AGGTYPE_ABS_SUM:
+            case AGGTYPE_GMV:
             case AGGTYPE_MUL:
             case AGGTYPE_DISTINCT_COUNT:
             case AGGTYPE_DISTINCT_LEAF:

rust/perspective-server/cpp/perspective/src/cpp/extract_aggregate.cpp

@@ -54,6 +54,7 @@ extract_aggregate(
         case AGGTYPE_SUM:
         case AGGTYPE_SUM_ABS:
         case AGGTYPE_ABS_SUM:
+        case AGGTYPE_GMV:
         case AGGTYPE_SUM_NOT_NULL:
         case AGGTYPE_MUL:
         case AGGTYPE_COUNT:

rust/perspective-server/cpp/perspective/src/cpp/server.cpp

@@ -1453,6 +1453,7 @@ ProtoServer::_handle_request(std::uint32_t client_id, Request&& req) {
             number_opts.add_aggregates()->set_name("distinct count");
             number_opts.add_aggregates()->set_name("dominant");
             number_opts.add_aggregates()->set_name("first");
+            number_opts.add_aggregates()->set_name("gmv");
             number_opts.add_aggregates()->set_name("high");
             number_opts.add_aggregates()->set_name("low");
             number_opts.add_aggregates()->set_name("max");

rust/perspective-server/cpp/perspective/src/cpp/sparse_tree.cpp

@@ -1815,6 +1815,67 @@ t_stree::update_agg_table(
                 );
                 dst->set_scalar(dst_ridx, new_value);
             } break;
+            case AGGTYPE_GMV: {
+                // Leaf nodes: plain `sum` of the underlying rows.
+                // Non-leaf nodes: sum over immediate children of
+                // `abs(sum(child_subtree))`. The child sums are recomputed
+                // from the gstate rather than read from already-populated
+                // child rows, so the parent does not see the abs-rolled-up
+                // child value — it sees the child's raw signed sum.
+                old_value.set(dst->get_scalar(dst_ridx));
+                const auto& col_name = spec.get_dependencies()[0].name();
+                auto dst_dtype = dst->get_dtype();
+                auto sum_reducer = [dst_dtype](std::vector<t_tscalar>& values
+                                   ) -> t_tscalar {
+                    if (values.empty()) {
+                        return mknone();
+                    }
+                    t_tscalar v;
+                    v.set(std::uint64_t(0));
+                    v.m_type = dst_dtype;
+                    for (const auto& x : values) {
+                        if (x.is_nan()) {
+                            continue;
+                        }
+                        v = v.add(x.coerce_numeric_dtype(dst_dtype));
+                    }
+                    return v;
+                };
+
+                if (is_leaf(nidx)) {
+                    auto pkeys = get_pkeys(nidx);
+                    new_value.set(
+                        reduce_from_gstate<std::function<
+                            t_tscalar(std::vector<t_tscalar>&)>>(
+                            gstate,
+                            expression_master_table,
+                            col_name,
+                            pkeys,
+                            sum_reducer
+                        )
+                    );
+                } else {
+                    t_tscalar rval;
+                    rval.set(std::uint64_t(0));
+                    rval.m_type = dst_dtype;
+                    for (auto cidx : get_child_idx(nidx)) {
+                        auto cpkeys = get_pkeys(cidx);
+                        t_tscalar csum = reduce_from_gstate<std::function<
+                            t_tscalar(std::vector<t_tscalar>&)>>(
+                            gstate,
+                            expression_master_table,
+                            col_name,
+                            cpkeys,
+                            sum_reducer
+                        );
+                        if (csum.is_valid() && !csum.is_nan()) {
+                            rval = rval.add(csum.abs());
+                        }
+                    }
+                    new_value.set(rval);
+                }
+                dst->set_scalar(dst_ridx, new_value);
+            } break;
             case AGGTYPE_MUL: {
                 old_value.set(dst->get_scalar(dst_ridx));
                 auto pkeys = get_pkeys(nidx);

rust/perspective-server/cpp/perspective/src/include/perspective/base.h

@@ -293,7 +293,8 @@ enum t_aggtype {
     AGGTYPE_PCT_SUM_PARENT,
     AGGTYPE_PCT_SUM_GRAND_TOTAL,
     AGGTYPE_VARIANCE,
-    AGGTYPE_STANDARD_DEVIATION
+    AGGTYPE_STANDARD_DEVIATION,
+    AGGTYPE_GMV
 };
 
 PERSPECTIVE_EXPORT t_aggtype str_to_aggtype(const std::string& str);