From 8dff3f1c6bb46e57dd90dcd3c802bd7982d56616 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 25 Apr 2026 14:59:03 +0000
Subject: [PATCH 1/3] Initial plan
From eec09812be59db225043a99b86b8cd1840940a25 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Sat, 25 Apr 2026 15:09:34 +0000
Subject: [PATCH 2/3] Iteration 281: extract strGetDummies into dedicated
module
Agent-Logs-Url: https://github.com/githubnext/tsessebe/sessions/615193c1-c4f0-473c-bd90-e155a60846a7
Co-authored-by: mrjf <180956+mrjf@users.noreply.github.com>
---
playground/str_get_dummies.html | 109 ++++++++++
src/index.ts | 3 +-
src/stats/index.ts | 3 +-
src/stats/str_get_dummies.ts | 129 ++++++++++++
src/stats/string_ops.ts | 88 +-------
tests/stats/str_get_dummies.test.ts | 304 ++++++++++++++++++++++++++++
tests/stats/string_ops.test.ts | 65 +-----
7 files changed, 548 insertions(+), 153 deletions(-)
create mode 100644 playground/str_get_dummies.html
create mode 100644 src/stats/str_get_dummies.ts
create mode 100644 tests/stats/str_get_dummies.test.ts
diff --git a/playground/str_get_dummies.html b/playground/str_get_dummies.html
new file mode 100644
index 00000000..467403eb
--- /dev/null
+++ b/playground/str_get_dummies.html
@@ -0,0 +1,109 @@
+
+
+
+ tsb — str.get_dummies: multi-label string encoding
+
+
+
+ str.get_dummies — multi-label string encoding
+
+ Port of pandas.Series.str.get_dummies(sep). Splits each
+ string by a separator (default "|") and returns a
+ DataFrame of binary indicator columns — one per unique token,
+ sorted lexicographically. null / undefined /
+ NaN values produce a row of all zeros.
+
+ ← back to index
+
+
+
Example 1 — basic split on |
+
+
Run
+
(click Run)
+
+
Example 2 — custom separator
+
+
Run
+
(click Run)
+
+
Example 3 — null / undefined / NaN → all-zero rows
+
+
Run
+
(click Run)
+
+
Example 4 — preserved Series index
+
+
Run
+
(click Run)
+
,
+ options: StrGetDummiesOptions = {},
+): DataFrame {
+ const sep = options.sep ?? "|";
+ const prefix = options.prefix ?? "";
+ const prefixSep = options.prefixSep ?? "_";
+ const colName = (token: string): string =>
+ prefix === "" ? token : `${prefix}${prefixSep}${token}`;
+ const vals = series.values;
+ const n = vals.length;
+
+ // Collect all unique tokens and per-row token sets.
+ const tokenSet = new Set();
+ const rowTokens: Set[] = new Array>(n);
+
+ for (let i = 0; i < n; i++) {
+ const v = vals[i];
+ const tokens = new Set();
+ if (v !== null && v !== undefined && !(typeof v === "number" && Number.isNaN(v))) {
+ const str = typeof v === "string" ? v : String(v);
+ if (str !== "") {
+ for (const tok of str.split(sep)) {
+ tokens.add(tok);
+ tokenSet.add(tok);
+ }
+ }
+ }
+ rowTokens[i] = tokens;
+ }
+
+ // Sort tokens lexicographically (pandas sorts columns for get_dummies).
+ const columns = [...tokenSet].sort();
+
+ // Build one Series per column. Use a Map (rather than a plain object)
+ // so that lexicographic order is preserved even for integer-like token
+ // names (plain object keys re-order numeric strings).
+ const idx = series.index;
+ const colMap = new Map>();
+ for (const col of columns) {
+ const arr: Scalar[] = new Array(n);
+ for (let i = 0; i < n; i++) {
+ arr[i] = rowTokens[i]?.has(col) === true ? 1 : 0;
+ }
+ colMap.set(colName(col), new Series({ data: arr, index: idx }));
+ }
+
+ return new DataFrame(colMap, idx);
+}
diff --git a/src/stats/string_ops.ts b/src/stats/string_ops.ts
index b5b486fb..5a444c50 100644
--- a/src/stats/string_ops.ts
+++ b/src/stats/string_ops.ts
@@ -10,7 +10,6 @@
* the accessor or better expressed as pure standalone utilities:
*
* - `strNormalize` — Unicode normalization (NFC / NFD / NFKC / NFKD)
- * - `strGetDummies` — split strings by delimiter → one-hot DataFrame
* - `strExtractAll` — extract ALL regex matches per element
* - `strRemovePrefix` — remove a leading prefix
* - `strRemoveSuffix` — remove a trailing suffix
@@ -21,7 +20,7 @@
* @module
*/
-import { DataFrame, Series } from "../core/index.ts";
+import { Series } from "../core/index.ts";
import type { Scalar } from "../types.ts";
// ─── public types ─────────────────────────────────────────────────────────────
@@ -32,27 +31,6 @@ export type NormalizeForm = "NFC" | "NFD" | "NFKC" | "NFKD";
/** Input accepted by all string-op functions. */
export type StrInput = Series | readonly Scalar[] | readonly string[] | string;
-/** Options for {@link strGetDummies}. */
-export interface StrGetDummiesOptions {
- /**
- * The delimiter used to split each element into tokens.
- * @default "|"
- */
- readonly sep?: string;
-
- /**
- * Prefix prepended to every column name in the output DataFrame.
- * @default ""
- */
- readonly prefix?: string;
-
- /**
- * Separator between the prefix and the token name.
- * @default "_"
- */
- readonly prefixSep?: string;
-}
-
/** Options for {@link strExtractAll}. */
export interface ExtractAllOptions {
/**
@@ -138,70 +116,6 @@ export function strNormalize(
return buildSeries(data, input);
}
-// ─── strGetDummies ────────────────────────────────────────────────────────────
-
-/**
- * Encode each string element as a row in a one-hot DataFrame by splitting on a
- * delimiter.
- *
- * Mirrors `pandas.Series.str.get_dummies(sep)`.
- *
- * @param input - Series or string array.
- * @param options - `sep` (default `"|"`), `prefix` and `prefixSep` for column names.
- * @returns A `DataFrame` of 0/1 integer values, one column per unique token.
- *
- * @example
- * ```ts
- * const s = new Series({ data: ["a|b", "b|c", "a"] });
- * strGetDummies(s);
- * // DataFrame
- * // a b c
- * // 0 1 1 0
- * // 1 0 1 1
- * // 2 1 0 0
- * ```
- */
-export function strGetDummies(
- input: readonly string[] | Series,
- options: StrGetDummiesOptions = {},
-): DataFrame {
- const sep = options.sep ?? "|";
- const prefix = options.prefix ?? "";
- const prefixSep = options.prefixSep ?? "_";
-
- const strs = toStringArray(input);
-
- // 1. Collect all unique tokens in first-seen order.
- const seen = new Set();
- const tokenRows: string[][] = strs.map((s) => {
- const tokens = s === "" ? [] : s.split(sep);
- for (const t of tokens) {
- seen.add(t);
- }
- return tokens;
- });
-
- const allTokens = [...seen].sort(); // stable alphabetical order
-
- // 2. Build column name with optional prefix.
- const colName = (token: string): string =>
- prefix === "" ? token : `${prefix}${prefixSep}${token}`;
-
- // 3. Build one Scalar[] per column.
- const columns: Record = {};
- for (const token of allTokens) {
- const name = colName(token);
- columns[name] = tokenRows.map((row) => (row.includes(token) ? 1 : 0));
- }
-
- // 4. Preserve the row index from a Series input.
- if (input instanceof Series) {
- const rowIndex = input.index;
- return DataFrame.fromColumns(columns, { index: rowIndex });
- }
- return DataFrame.fromColumns(columns);
-}
-
// ─── strExtractAll ────────────────────────────────────────────────────────────
/**
diff --git a/tests/stats/str_get_dummies.test.ts b/tests/stats/str_get_dummies.test.ts
new file mode 100644
index 00000000..30f39718
--- /dev/null
+++ b/tests/stats/str_get_dummies.test.ts
@@ -0,0 +1,304 @@
+/**
+ * Tests for strGetDummies — pandas.Series.str.get_dummies(sep) port.
+ */
+
+import { describe, expect, test } from "bun:test";
+import * as fc from "fast-check";
+import { DataFrame, Series, strGetDummies } from "../../src/index.ts";
+import type { Scalar } from "../../src/index.ts";
+
+function s(data: readonly Scalar[]): Series {
+ return new Series({ data: [...data] });
+}
+
+// ─── Basic functionality ──────────────────────────────────────────────────────
+
+describe("strGetDummies — basic", () => {
+ test("splits by default pipe separator", () => {
+ const df = strGetDummies(s(["a|b", "b|c", "a"]));
+ expect(df).toBeInstanceOf(DataFrame);
+ expect([...df.columns.values]).toEqual(["a", "b", "c"]);
+ expect(df.shape).toEqual([3, 3]);
+ });
+
+ test("correct indicator values", () => {
+ const df = strGetDummies(s(["a|b", "b|c", "a"]));
+ expect([...df.col("a").values]).toEqual([1, 0, 1]);
+ expect([...df.col("b").values]).toEqual([1, 1, 0]);
+ expect([...df.col("c").values]).toEqual([0, 1, 0]);
+ });
+
+ test("custom separator", () => {
+ const df = strGetDummies(s(["x,y", "y,z", "x"]), { sep: "," });
+ expect([...df.columns.values]).toEqual(["x", "y", "z"]);
+ expect([...df.col("x").values]).toEqual([1, 0, 1]);
+ });
+
+ test("single-value elements", () => {
+ const df = strGetDummies(s(["a", "b", "c"]));
+ expect([...df.columns.values]).toEqual(["a", "b", "c"]);
+ expect([...df.col("a").values]).toEqual([1, 0, 0]);
+ });
+
+ test("columns are sorted lexicographically", () => {
+ const df = strGetDummies(s(["z|a|m", "a|z"]));
+ expect([...df.columns.values]).toEqual(["a", "m", "z"]);
+ });
+
+ test("null values produce all-zero rows", () => {
+ const df = strGetDummies(s(["a|b", null, "b"]));
+ expect([...df.col("a").values]).toEqual([1, 0, 0]);
+ expect([...df.col("b").values]).toEqual([1, 0, 1]);
+ });
+
+ test("undefined values produce all-zero rows", () => {
+ const df = strGetDummies(s(["a", undefined as unknown as Scalar, "b"]));
+ expect([...df.col("a").values]).toEqual([1, 0, 0]);
+ expect([...df.col("b").values]).toEqual([0, 0, 1]);
+ });
+
+ test("NaN values produce all-zero rows", () => {
+ const df = strGetDummies(s(["a|b", Number.NaN, "b"]));
+ expect([...df.col("a").values]).toEqual([1, 0, 0]);
+ expect([...df.col("b").values]).toEqual([1, 0, 1]);
+ });
+
+ test("empty-string elements produce all-zero rows", () => {
+ const df = strGetDummies(s(["a|b", ""]));
+ expect([...df.col("a").values]).toEqual([1, 0]);
+ expect([...df.col("b").values]).toEqual([1, 0]);
+ });
+
+ test("all nulls returns empty-column DataFrame with preserved row count", () => {
+ const df = strGetDummies(s([null, null]));
+ expect([...df.columns.values]).toEqual([]);
+ expect(df.shape).toEqual([2, 0]);
+ });
+
+ test("empty series returns empty DataFrame", () => {
+ const df = strGetDummies(s([]));
+ expect(df.shape).toEqual([0, 0]);
+ });
+
+ test("preserves original index", () => {
+ const ser = new Series({ data: ["a|b", "c"], index: ["r1", "r2"] });
+ const df = strGetDummies(ser);
+ expect([...df.index.values]).toEqual(["r1", "r2"]);
+ });
+
+ test("duplicate tokens in same element are counted once", () => {
+ const df = strGetDummies(s(["a|a|b"]));
+ expect([...df.col("a").values]).toEqual([1]);
+ expect([...df.col("b").values]).toEqual([1]);
+ });
+
+ test("whitespace tokens are preserved as-is (not stripped)", () => {
+ const df = strGetDummies(s([" a | b "]));
+ expect([...df.columns.values]).toEqual([" a ", " b "]);
+ });
+
+ test("prefix and prefixSep option", () => {
+ const df = strGetDummies(s(["x|y"]), { prefix: "tag", prefixSep: "-" });
+ expect([...df.columns.values]).toEqual(["tag-x", "tag-y"]);
+ });
+});
+
+// ─── Result shape and types ───────────────────────────────────────────────────
+
+describe("strGetDummies — result shape", () => {
+ test("row count matches input series length", () => {
+ const df = strGetDummies(s(["a|b", "c", "d|e|f"]));
+ expect(df.shape[0]).toBe(3);
+ });
+
+ test("column count equals unique token count", () => {
+ const df = strGetDummies(s(["a|b|c", "b|d"]));
+ expect(df.shape[1]).toBe(4);
+ });
+
+ test("all values are 0 or 1", () => {
+ const df = strGetDummies(s(["x|y", "y|z", "x|z"]));
+ for (const col of df.columns.values as readonly string[]) {
+ for (const v of df.col(col).values) {
+ expect(v === 0 || v === 1).toBe(true);
+ }
+ }
+ });
+
+ test("sum of row values equals unique-token count per row", () => {
+ const df = strGetDummies(s(["a|b|c", "a", "b|c"]));
+ const cols = df.columns.values as readonly string[];
+ const rowSums = [0, 1, 2].map((i) =>
+ cols.reduce((acc, col) => {
+ const v = df.col(col).values[i];
+ return acc + (typeof v === "number" ? v : 0);
+ }, 0),
+ );
+ expect(rowSums).toEqual([3, 1, 2]);
+ });
+});
+
+// ─── Pandas parity examples ───────────────────────────────────────────────────
+
+describe("strGetDummies — pandas parity", () => {
+ test("pandas example: a|b, b|c, a", () => {
+ // >>> pd.Series(['a|b', 'b|c', 'a']).str.get_dummies()
+ // a b c
+ // 0 1 1 0
+ // 1 0 1 1
+ // 2 1 0 0
+ const df = strGetDummies(s(["a|b", "b|c", "a"]));
+ expect(df.toRecords()).toEqual([
+ { a: 1, b: 1, c: 0 },
+ { a: 0, b: 1, c: 1 },
+ { a: 1, b: 0, c: 0 },
+ ]);
+ });
+
+ test("pandas example: custom sep comma", () => {
+ const df = strGetDummies(s(["a,b", "b,c", "a"]), { sep: "," });
+ expect(df.toRecords()).toEqual([
+ { a: 1, b: 1, c: 0 },
+ { a: 0, b: 1, c: 1 },
+ { a: 1, b: 0, c: 0 },
+ ]);
+ });
+
+ test("multi-label tags scenario", () => {
+ const df = strGetDummies(
+ s(["python|pandas", "python|numpy", "pandas|numpy|scipy"]),
+ );
+ expect([...df.columns.values]).toEqual(["numpy", "pandas", "python", "scipy"]);
+ expect([...df.col("python").values]).toEqual([1, 1, 0]);
+ expect([...df.col("scipy").values]).toEqual([0, 0, 1]);
+ });
+});
+
+// ─── Property-based tests ─────────────────────────────────────────────────────
+
+describe("strGetDummies — property-based", () => {
+ test("row count always equals series length", () => {
+ fc.assert(
+ fc.property(
+ fc.array(fc.option(fc.string({ maxLength: 10 }), { nil: null }), {
+ minLength: 1,
+ maxLength: 20,
+ }),
+ (arr) => {
+ const df = strGetDummies(s(arr as Scalar[]));
+ return df.shape[0] === arr.length;
+ },
+ ),
+ );
+ });
+
+ test("all cell values are 0 or 1", () => {
+ fc.assert(
+ fc.property(
+ fc.array(
+ fc.option(
+ fc
+ .array(fc.string({ minLength: 1, maxLength: 5 }), {
+ minLength: 1,
+ maxLength: 4,
+ })
+ .map((parts) => parts.join("|")),
+ { nil: null },
+ ),
+ { minLength: 1, maxLength: 15 },
+ ),
+ (arr) => {
+ const df = strGetDummies(s(arr as Scalar[]));
+ for (const col of df.columns.values as readonly string[]) {
+ for (const v of df.col(col).values) {
+ if (v !== 0 && v !== 1) {
+ return false;
+ }
+ }
+ }
+ return true;
+ },
+ ),
+ );
+ });
+
+ test("index is preserved", () => {
+ fc.assert(
+ fc.property(
+ fc.array(fc.string({ maxLength: 10 }), { minLength: 1, maxLength: 15 }),
+ (arr) => {
+ const ser = s(arr as Scalar[]);
+ const df = strGetDummies(ser);
+ const origIdx = [...ser.index.values];
+ const dfIdx = [...df.index.values];
+ return (
+ origIdx.length === dfIdx.length && origIdx.every((v, i) => v === dfIdx[i])
+ );
+ },
+ ),
+ );
+ });
+
+ test("columns are always sorted lexicographically", () => {
+ fc.assert(
+ fc.property(
+ fc.array(
+ fc
+ .array(fc.string({ minLength: 1, maxLength: 5 }), {
+ minLength: 1,
+ maxLength: 4,
+ })
+ .map((parts) => parts.join("|")),
+ { minLength: 1, maxLength: 10 },
+ ),
+ (arr) => {
+ const df = strGetDummies(s(arr as Scalar[]));
+ const cols = (df.columns.values as readonly string[]).map(String);
+ const sorted = [...cols].sort();
+ return cols.every((c, i) => c === sorted[i]);
+ },
+ ),
+ );
+ });
+
+ test("null/undefined/NaN rows always have row-sum 0", () => {
+ fc.assert(
+ fc.property(
+ fc.array(
+ fc.oneof(
+ fc.constant(null),
+ fc.constant(undefined),
+ fc.constant(Number.NaN),
+ fc
+ .array(fc.string({ minLength: 1, maxLength: 4 }), {
+ minLength: 1,
+ maxLength: 3,
+ })
+ .map((parts) => parts.join("|")),
+ ),
+ { minLength: 1, maxLength: 10 },
+ ),
+ (arr) => {
+ const df = strGetDummies(s(arr as Scalar[]));
+ const cols = df.columns.values as readonly string[];
+ for (let i = 0; i < arr.length; i++) {
+ const v = arr[i];
+ const isMissing =
+ v === null ||
+ v === undefined ||
+ (typeof v === "number" && Number.isNaN(v));
+ if (!isMissing) {
+ continue;
+ }
+ for (const col of cols) {
+ if (df.col(col).values[i] !== 0) {
+ return false;
+ }
+ }
+ }
+ return true;
+ },
+ ),
+ );
+ });
+});
diff --git a/tests/stats/string_ops.test.ts b/tests/stats/string_ops.test.ts
index 0f5b84cc..435af513 100644
--- a/tests/stats/string_ops.test.ts
+++ b/tests/stats/string_ops.test.ts
@@ -1,6 +1,6 @@
/**
* Tests for src/stats/string_ops.ts
- * — strNormalize, strGetDummies, strExtractAll, strRemovePrefix,
+ * — strNormalize, strExtractAll, strRemovePrefix,
* strRemoveSuffix, strTranslate, strCharWidth, strByteLength
*/
import { describe, expect, it } from "bun:test";
@@ -11,7 +11,6 @@ import {
strByteLength,
strCharWidth,
strExtractAll,
- strGetDummies,
strNormalize,
strRemovePrefix,
strRemoveSuffix,
@@ -78,68 +77,6 @@ describe("strNormalize", () => {
});
});
-// ─── strGetDummies ────────────────────────────────────────────────────────────
-
-describe("strGetDummies", () => {
- it("basic | separator", () => {
- const df = strGetDummies(s(["a|b", "b|c", "a"]));
- expect(df.shape[0]).toBe(3);
- expect([...df.columns.values].sort()).toEqual(["a", "b", "c"]);
- expect(df.col("a").values[0]).toBe(1);
- expect(df.col("a").values[1]).toBe(0);
- expect(df.col("a").values[2]).toBe(1);
- expect(df.col("b").values[0]).toBe(1);
- expect(df.col("b").values[1]).toBe(1);
- expect(df.col("b").values[2]).toBe(0);
- expect(df.col("c").values[0]).toBe(0);
- expect(df.col("c").values[1]).toBe(1);
- expect(df.col("c").values[2]).toBe(0);
- });
-
- it("custom separator", () => {
- const df = strGetDummies(s(["a,b", "b,c"]), { sep: "," });
- expect([...df.columns.values].sort()).toEqual(["a", "b", "c"]);
- });
-
- it("prefix option", () => {
- const df = strGetDummies(s(["x|y"]), { prefix: "tag", prefixSep: "-" });
- expect([...df.columns.values].sort()).toEqual(["tag-x", "tag-y"]);
- });
-
- it("empty string element maps to no tokens", () => {
- const df = strGetDummies(s(["a|b", ""]));
- expect(df.col("a").values[1]).toBe(0);
- expect(df.col("b").values[1]).toBe(0);
- });
-
- it("single-token element", () => {
- const df = strGetDummies(s(["a", "b", "a"]));
- expect(df.shape[0]).toBe(3);
- expect(df.col("a").values[0]).toBe(1);
- expect(df.col("a").values[1]).toBe(0);
- expect(df.col("a").values[2]).toBe(1);
- });
-
- it("all same token → single column of ones", () => {
- const df = strGetDummies(s(["x", "x", "x"]));
- expect(df.shape[1]).toBe(1);
- expect([...df.col("x").values]).toEqual([1, 1, 1]);
- });
-
- it("preserves Series index in output rows", () => {
- const ser = new Series({ data: ["a|b", "b"] as Scalar[], index: [10, 20] });
- const df = strGetDummies(ser);
- expect(df.index.values[0]).toBe(10);
- expect(df.index.values[1]).toBe(20);
- });
-
- it("array input (not Series)", () => {
- const df = strGetDummies(["a|b", "c"]);
- expect(df.shape[0]).toBe(2);
- expect([...df.columns.values].sort()).toEqual(["a", "b", "c"]);
- });
-});
-
// ─── strExtractAll ────────────────────────────────────────────────────────────
describe("strExtractAll", () => {
From 917c7e69a485cd11d6c89b09fa2f59ad59d600a9 Mon Sep 17 00:00:00 2001
From: "github-actions[bot]"
<41898282+github-actions[bot]@users.noreply.github.com>
Date: Sat, 25 Apr 2026 19:14:18 +0000
Subject: [PATCH 3/3] fix: apply biome formatting to str_get_dummies test
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---
tests/stats/str_get_dummies.test.ts | 12 +++---------
1 file changed, 3 insertions(+), 9 deletions(-)
diff --git a/tests/stats/str_get_dummies.test.ts b/tests/stats/str_get_dummies.test.ts
index 30f39718..924f2c1e 100644
--- a/tests/stats/str_get_dummies.test.ts
+++ b/tests/stats/str_get_dummies.test.ts
@@ -165,9 +165,7 @@ describe("strGetDummies — pandas parity", () => {
});
test("multi-label tags scenario", () => {
- const df = strGetDummies(
- s(["python|pandas", "python|numpy", "pandas|numpy|scipy"]),
- );
+ const df = strGetDummies(s(["python|pandas", "python|numpy", "pandas|numpy|scipy"]));
expect([...df.columns.values]).toEqual(["numpy", "pandas", "python", "scipy"]);
expect([...df.col("python").values]).toEqual([1, 1, 0]);
expect([...df.col("scipy").values]).toEqual([0, 0, 1]);
@@ -231,9 +229,7 @@ describe("strGetDummies — property-based", () => {
const df = strGetDummies(ser);
const origIdx = [...ser.index.values];
const dfIdx = [...df.index.values];
- return (
- origIdx.length === dfIdx.length && origIdx.every((v, i) => v === dfIdx[i])
- );
+ return origIdx.length === dfIdx.length && origIdx.every((v, i) => v === dfIdx[i]);
},
),
);
@@ -284,9 +280,7 @@ describe("strGetDummies — property-based", () => {
for (let i = 0; i < arr.length; i++) {
const v = arr[i];
const isMissing =
- v === null ||
- v === undefined ||
- (typeof v === "number" && Number.isNaN(v));
+ v === null || v === undefined || (typeof v === "number" && Number.isNaN(v));
if (!isMissing) {
continue;
}