Adding DF upsample (#163)

* Adding DF upsample * Update polars/dataframe.ts Co-authored-by: universalmind303 <[email protected]> * Adding semi * Updating packages Simplifying test Fixing documentation * Replacing pl.all with pl.col(*) * Updating yarn lock file * Removing toString from DF test --------- Co-authored-by: universalmind303 <[email protected]>
pola-rs · Feb 27, 2024 · 8ecf9bd · 8ecf9bd
1 parent 208000f
commit 8ecf9bd
Show file tree

Hide file tree

Showing 5 changed files with 376 additions and 195 deletions.
diff --git a/__tests__/dataframe.test.ts b/__tests__/dataframe.test.ts
@@ -2249,7 +2249,6 @@ describe("meta", () => {
     expect(dfString).toStrictEqual(expected);
   });
 });
-
 test("Jupyter.display", () => {
   const df = pl.DataFrame({
     os: ["apple", "linux"],
@@ -2329,4 +2328,79 @@ describe("additional", () => {
     const actual = df.toRecords();
     expect(actual).toEqual(rows);
   });
+  test("upsample", () => {
+    const df = pl
+      .DataFrame({
+        date: [
+          new Date(2024, 1, 1),
+          new Date(2024, 3, 1),
+          new Date(2024, 4, 1),
+          new Date(2024, 5, 1),
+        ],
+        groups: ["A", "B", "A", "B"],
+        values: [0, 1, 2, 3],
+      })
+      .withColumn(pl.col("date").cast(pl.Date).alias("date"))
+      .sort("date");
+
+    let actual = df
+      .upsample("date", "1mo", "0ns", "groups", true)
+      .select(pl.col("*").forwardFill());
+
+    let expected = pl
+      .DataFrame({
+        date: [
+          new Date(2024, 1, 1),
+          new Date(2024, 2, 1),
+          new Date(2024, 3, 1),
+          new Date(2024, 4, 1),
+          new Date(2024, 3, 1),
+          new Date(2024, 4, 1),
+          new Date(2024, 5, 1),
+        ],
+        groups: ["A", "A", "A", "A", "B", "B", "B"],
+        values: [0.0, 0.0, 0.0, 2.0, 1.0, 1.0, 3.0],
+      })
+      .withColumn(pl.col("date").cast(pl.Date).alias("date"));
+
+    expect(actual).toFrameEqual(expected);
+
+    actual = df
+      .upsample({
+        timeColumn: "date",
+        every: "1mo",
+        offset: "0ns",
+        by: "groups",
+        maintainOrder: true,
+      })
+      .select(pl.col("*").forwardFill());
+
+    expect(actual).toFrameEqual(expected);
+
+    actual = df
+      .upsample({ timeColumn: "date", every: "1mo" })
+      .select(pl.col("*").forwardFill());
+
+    expected = pl
+      .DataFrame({
+        date: [
+          new Date(2024, 1, 1),
+          new Date(2024, 2, 1),
+          new Date(2024, 3, 1),
+          new Date(2024, 4, 1),
+          new Date(2024, 5, 1),
+        ],
+        groups: ["A", "A", "B", "A", "B"],
+        values: [0.0, 0.0, 1.0, 2.0, 3.0],
+      })
+      .withColumn(pl.col("date").cast(pl.Date).alias("date"));
+
+    expect(actual).toFrameEqual(expected);
+
+    actual = df
+      .upsample({ timeColumn: "date", every: "1m" })
+      .select(pl.col("*").forwardFill());
+
+    expect(actual.shape).toEqual({ height: 174_241, width: 3 });
+  });
 });
diff --git a/package.json b/package.json
@@ -54,17 +54,17 @@
     "precommit": "yarn lint && yarn test"
   },
   "devDependencies": {
-    "@biomejs/biome": "^1.5.1",
-    "@napi-rs/cli": "^2.17.0",
+    "@biomejs/biome": "^1.5.3",
+    "@napi-rs/cli": "^2.18.0",
     "@types/chance": "^1.1.6",
-    "@types/jest": "^29.5.11",
-    "@types/node": "^20.10.8",
+    "@types/jest": "^29.5.12",
+    "@types/node": "^20.11.17",
     "chance": "^1.1.11",
     "jest": "^29.7.0",
     "source-map-support": "^0.5.21",
-    "ts-jest": "^29.1.1",
+    "ts-jest": "^29.1.2",
     "ts-node": "^10.9.2",
-    "typedoc": "^0.25.7",
+    "typedoc": "^0.25.8",
     "typescript": "5.3.3"
   },
   "packageManager": "[email protected]",