Fixed postgres floating point type (#952)

SilasMarvin · web-flow · commit eafc5b8c25f1 · 2023-08-25T12:15:14.000-07:00
diff --git a/pgml-sdks/rust/pgml/javascript/tests/typescript-tests/test.ts b/pgml-sdks/rust/pgml/javascript/tests/typescript-tests/test.ts
@@ -18,7 +18,9 @@ const generate_dummy_documents = (count: number) => {
     docs.push({
       id: i,
       text: `This is a test document: ${i}`,
+      project: "a10",
       uuid: i * 10,
+      floating_uuid: i * 1.1,
       name: `Test Document ${i}`,
     });
   }
@@ -36,7 +38,7 @@ it("can create collection", () => {
 
 it("can create model", () => {
   let model = pgml.newModel("test", "openai", {
-    "tester": "test 0123948712394871234987"
+    some_example_parameter: "test 0123948712394871234987",
   });
   expect(model).toBeTruthy();
 });
@@ -74,7 +76,7 @@ it("can vector search with local embeddings", async () => {
   await collection.archive();
 });
 
-it("can vector search with remote embeddings", async() => {
+it("can vector search with remote embeddings", async () => {
   let model = pgml.newModel("text-embedding-ada-002", "openai");
   let splitter = pgml.newSplitter();
   let pipeline = pgml.newPipeline("test_j_p_cvswre_0", model, splitter);
@@ -86,26 +88,34 @@ it("can vector search with remote embeddings", async() => {
   await collection.archive();
 });
 
-it("can vector search with query builder", async() => {
+it("can vector search with query builder", async () => {
   let model = pgml.newModel();
   let splitter = pgml.newSplitter();
   let pipeline = pgml.newPipeline("test_j_p_cvswqb_0", model, splitter);
   let collection = pgml.newCollection("test_j_c_cvswqb_1");
   await collection.upsert_documents(generate_dummy_documents(3));
   await collection.add_pipeline(pipeline);
-  let results = await collection.query().vector_recall("Here is some query", pipeline).limit(10).fetch_all();
+  let results = await collection
+    .query()
+    .vector_recall("Here is some query", pipeline)
+    .limit(10)
+    .fetch_all();
   expect(results).toHaveLength(3);
   await collection.archive();
 });
 
-it("can vector search with query builder with remote embeddings", async() => {
+it("can vector search with query builder with remote embeddings", async () => {
   let model = pgml.newModel("text-embedding-ada-002", "openai");
   let splitter = pgml.newSplitter();
   let pipeline = pgml.newPipeline("test_j_p_cvswqbwre_0", model, splitter);
   let collection = pgml.newCollection("test_j_c_cvswqbwre_1");
   await collection.upsert_documents(generate_dummy_documents(3));
   await collection.add_pipeline(pipeline);
-  let results = await collection.query().vector_recall("Here is some query", pipeline).limit(10).fetch_all();
+  let results = await collection
+    .query()
+    .vector_recall("Here is some query", pipeline)
+    .limit(10)
+    .fetch_all();
   expect(results).toHaveLength(3);
   await collection.archive();
 });
@@ -122,10 +132,12 @@ it("can vector search with query builder and metadata filtering", async () => {
     .vector_recall("Here is some query", pipeline)
     .filter({
       metadata: {
-        $or: [{ uuid: { $eq: 0 } }, { uuid: { $eq: 20 } }],
+        $or: [{ uuid: { $eq: 0 } }, { floating_uuid: { $lt: 2 } }],
+        project: { $eq: "a10" },
       },
     })
-    .limit(10).fetch_all();
+    .limit(10)
+    .fetch_all();
   expect(results).toHaveLength(2);
   await collection.archive();
 });
@@ -141,7 +153,6 @@ it("pipeline to dict", async () => {
   let collection = pgml.newCollection("test_j_c_ptd_2");
   await collection.add_pipeline(pipeline);
   let pipeline_dict = await pipeline.to_dict();
-  console.log(JSON.stringify(pipeline_dict))
   expect(pipeline_dict["name"]).toBe("test_j_p_ptd_0");
   await collection.archive();
 });
diff --git a/pgml-sdks/rust/pgml/python/tests/test.py b/pgml-sdks/rust/pgml/python/tests/test.py
@@ -29,7 +29,8 @@ def generate_dummy_documents(count: int) -> List[Dict[str, Any]]:
             {
                 "id": i,
                 "text": "This is a test document: {}".format(i),
-                "some_random_thing": "This will be metadata on it",
+                "project": "a10",
+                "floating_uuid": i * 1.01,
                 "uuid": i * 10,
                 "name": "Test Document {}".format(i),
             }
@@ -147,17 +148,18 @@ async def test_can_vector_search_with_query_builder_and_metadata_filtering():
     results = (
         await collection.query()
         .vector_recall("Here is some query", pipeline)
-        .filter({
-            "metadata": {
-                "uuid": {
-                    "$eq": 0
-                }
+        .filter(
+            {
+                "metadata": {
+                    "$or": [{"uuid": {"$eq": 0}}, {"floating_uuid": {"$lt": 2}}],
+                    "project": {"$eq": "a10"},
+                },
             }
-        })
+        )
         .limit(10)
         .fetch_all()
     )
-    assert len(results) == 1
+    assert len(results) == 2
     await collection.archive()
 
 
diff --git a/pgml-sdks/rust/pgml/src/filter_builder.rs b/pgml-sdks/rust/pgml/src/filter_builder.rs
@@ -117,9 +117,9 @@ fn get_value_type(value: &serde_json::Value) -> String {
     } else if value.is_string() {
         "text".to_string()
     } else if value.is_i64() {
-        "bigint".to_string()
+        "float8".to_string()
     } else if value.is_f64() {
-        "double".to_string()
+        "float8".to_string()
     } else if value.is_boolean() {
         "bool".to_string()
     } else {
