{ "category": "Lookup and reference functions", "display_name": "HSTACK", "entry_kind": "built_in_function", "eval_api": { "method": "POST", "url": "https://handbook.dnacalc.com/api/eval" }, "function_id": "FUNC.HSTACK", "implementation_state": { "answer_bank": null, "detail": "Of 28 recorded answers (28 workbook cells), 5 do not match: 5 differ in kind or value.", "dnacalc_commit": "9e1884369a678158dd82d3d74188daf986995e87", "in_catalogue": true, "state": "differs", "state_label": "Known larger differences", "summary": "Known larger differences on 5 of 28 recorded answers", "workbook_corpus": { "cells": 28, "matched": 23, "other_mismatch": 5 } }, "implementations": [ { "commit": "9e1884369a678158dd82d3d74188daf986995e87", "flavour": "excel-compatibility-candidate", "language": "Rust", "name": "CalcFunc", "own_code": [ "https://github.com/DnaCalc/DnaCalc/blob/9e1884369a678158dd82d3d74188daf986995e87/crates/calc-func/src/catalogue/arrays.rs#L671-L682", "https://github.com/DnaCalc/DnaCalc/blob/9e1884369a678158dd82d3d74188daf986995e87/crates/calc-func/src/ported/dynamic_array_reshape_family.rs#L1126-L1145" ], "repository": "https://github.com/DnaCalc/DnaCalc", "shared_code": [ "catalogue/arrays::prepared", "catalogue/arrays::prepared_one", "catalogue/arrays::result", "catalogue/arrays::run", "ported/dynamic_array_reshape_family::shape", "ported/dynamic_array_reshape_family::get", "ported/dynamic_array_reshape_family::build_array", "ported/dynamic_array_reshape_family::map_dynamic_array_reshape_error_to_ws" ], "source": "https://github.com/DnaCalc/DnaCalc/blob/9e1884369a678158dd82d3d74188daf986995e87/crates/calc-func/src/catalogue/arrays.rs#L671-L682", "state": "differs", "suite": null, "summary": "Known larger differences on 5 of 28 recorded answers" } ], "json_url": "https://handbook.dnacalc.com/api/functions/FUNC.HSTACK.json", "name": "HSTACK", "notes": [ { "basis": "27 recorded readings in dynamic-array Excel, each with size observers.", "documentation": null, "evidence": [ { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "scalar_errors_keep_their_array_argument_origins", "note": "recorded readings pinned", "path": "apps/dna-onecalc/tests/acceptance.rs", "repository": "DnaCalc" }, { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheets fn2_q001 to fn2_q007, fn2_q031 to fn2_q042, fn2_q232 to fn2_q239", "note": "recorded readings", "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } ], "functions": [ "FUNC.BYCOL", "FUNC.BYROW", "FUNC.EXPAND", "FUNC.HSTACK", "FUNC.MAP", "FUNC.TAKE", "FUNC.TOCOL", "FUNC.TOROW", "FUNC.VSTACK" ], "inferred": null, "kind": "quirk", "note_id": "FN-HSTACK-001", "open": [ "Competing errors in later positions have not been read in a published capture." ], "recorded_on": "2026-11-04", "repros": [ { "call": "=HSTACK(1,#VALUE!)", "correct": null, "excel": { "bits": null, "display": "#VALUE!", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q004, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=TAKE(#N/A,0)", "correct": null, "excel": { "bits": null, "display": "#N/A", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q232, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=MAP(NA(),LAMBDA(xx,IFERROR(xx,3)))", "correct": null, "excel": { "bits": null, "display": "#N/A", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q031, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=MAP({#N/A},LAMBDA(xx,3))", "correct": null, "excel": { "bits": "0x4008000000000000", "display": "3", "kind": "number" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q032, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=HSTACK(Data!C1,{1;2})", "correct": null, "excel": { "bits": null, "display": "{#N/A,1;#N/A,2}", "kind": "array" }, "input_route": "cell-value2", "inputs": [ { "name": "Data!C1", "value": { "bits": null, "display": "#N/A", "kind": "error" } } ], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q238, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=TOCOL(#N/A,2)", "correct": null, "excel": { "bits": null, "display": "#N/A", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q233, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=TOCOL({#N/A},2)", "correct": null, "excel": { "bits": null, "display": "#CALC!", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q234, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=EXPAND(Data!C1,2,2,\"pad\")", "correct": null, "excel": { "bits": null, "display": "#N/A", "kind": "error" }, "input_route": "cell-value2", "inputs": [ { "name": "Data!C1", "value": { "bits": null, "display": "#N/A", "kind": "error" } } ], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q237, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=BYROW(#N/A,#VALUE!)", "correct": null, "excel": { "bits": null, "display": "#N/A", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q040, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } }, { "call": "=BYROW({#N/A},2)", "correct": null, "excel": { "bits": null, "display": "#VALUE!", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "source": { "commit": "2d0eddcac86413014643af81d646581d54ba57bc", "locator": "sheet fn2_q038, row 20, column 10", "note": null, "path": "crates/calc-func/tests/captures/fn2_questions/oracle.capture.json", "repository": "DnaCalc" } } ], "schema": "efh.function-note/v1", "statement": "An error written directly as an argument of these functions is the answer for the whole call. It is not placed in the result: HSTACK(1,#VALUE!) is #VALUE!, not a row holding 1 and an error, and TAKE(#N/A,0) is #N/A. With MAP the LAMBDA does not run at all: MAP(NA(),LAMBDA(xx,IFERROR(xx,3))) is #N/A. The same error inside an array, or in a referenced cell, is an ordinary element: MAP({#N/A},LAMBDA(xx,3)) is 3, and HSTACK(Data!C1,{1;2}) with #N/A in Data!C1 is a 2 by 2 array with #N/A in its first column. TOCOL and TOROW show the difference with ignore 2: TOCOL(#N/A,2) is #N/A, while TOCOL({#N/A},2) drops the error, has nothing left, and is #CALC!. EXPAND is an exception on the cell side: EXPAND(Data!C1,2,2,\"pad\") is #N/A for the whole call. BYROW and BYCOL check the direct error first too: BYROW(#N/A,#VALUE!) is #N/A, while BYROW({#N/A},2), where 2 is not a LAMBDA, is #VALUE!.", "status": "measured", "title": "A direct error stops these array functions; the same error in an array or cell is an element" } ], "schema": "efh.function-reference/v1", "section_order": [ "what-it-does", "try-it", "quirks", "how-to-implement", "version-differences", "alternatives" ], "sections": { "alternatives": "\n### Relationships\n\n\n- **`VSTACK`** stacks vertically: widths are padded instead of heights, and the result's height\n is the sum. The pair is symmetric, and `HSTACK(VSTACK(…), …)` compositions are how\n dynamic-array formulas assemble tables.\n- **`EXPAND`** is how you control the padding rather than accept `#N/A`: expand each piece to the\n common height with your own pad value first, then stack.\n- **`CHOOSECOLS`** is the inverse in spirit — it takes columns out of one array where `HSTACK`\n puts columns from several together.\n- **`TOROW` / `TOCOL`** flatten rather than stack, and `WRAPROWS` / `WRAPCOLS` reshape a vector\n into a rectangle; all four answer the neighbouring question of \"how do I get this data into\n that shape\".\n- **The `&` operator and `TEXTJOIN`** concatenate *values*; `HSTACK` concatenates *arrays*. The\n word \"concatenate\" pulls readers toward the wrong one.\n- **`INDEX` / `OFFSET` gymnastics** were how workbooks assembled multi-source arrays before the\n stacking functions existed; `HSTACK` replaces most of that with a call whose intent is legible.\n\n\n", "how-to-implement": "\n### Notes for implementers\n\n\n- **The pad is per-argument, not per-result.** Each argument is padded to the common height\n independently; there is no single rectangle being filled. Implementations that build the\n output row-major must know each source's height to decide where padding begins in that row.\n- **Width is a sum and can overflow the grid.** Validate against the host's column limit before\n allocating, and decide what the failure looks like — the Handbook has not established Excel's\n behaviour at the boundary.\n- **Preserve element kinds exactly.** Stacking is a copy, not a conversion; text, logicals and\n errors must arrive unchanged. This is the property that lets `HSTACK` build heterogeneous\n tables, and the easiest one to lose in an implementation that types its buffers numerically.\n- **The identity case matters.** `HSTACK(x)` with a single argument is a legitimate call — the\n arity minimum is 1 — and it is used deliberately to materialize a reference as an array. It\n must not be optimized into a pass-through that preserves reference-ness.\n- This function has its own module in the reference implementation rather than sharing the\n reshape family's, which is a code-organization fact, not a behavioural one.\n\n\n", "quirks": "\n### Result and edge cases\n\n\nReturn kind: `Array`, spilled. Spilling is host-side adaptation\n([the call pipeline](../model/03-call-pipeline.md)).\n\n- **References are resolved first** (`ValuesOnlyPreAdapter`), so the result carries values, not\n reference structure. `HSTACK` cannot be used where a reference is required.\n- **A single argument** returns that argument as an array — the identity case, and a useful way\n to force a reference into array form.\n- **Element kinds are preserved**: no coercion is applied to the stacked elements, and errors\n inside the inputs pass through as values (`error_collapse_profile: None`).\n- **Ragged inputs are the normal case, not an error.** Padding is the documented behaviour;\n there is no strict mode.\n- **Whether a whole-column reference is admissible** as an argument, and what height it\n contributes, is not established here — it is the input most likely to interact badly with the\n \"height of the tallest argument\" rule.\n\n\n\n### Errors\n\n\nMicrosoft's page for `HSTACK` documents `#N/A` as the *pad value*, which is a value in the\nresult rather than a failure of the call. The page does not publish a general error table.\n`#VALUE!` for an inadmissible argument is the family's conventional code and is recorded here as\nexpected, not verified.\n\nAn error value supplied as an argument propagates under the universal rule in\n[coercion and lifting](../model/02-coercion-and-lifting.md) — as distinct from an error sitting\n*inside* an argument array, which is stacked like any other value.\n\nA result exceeding the grid's dimensions is a host limit rather than a function semantic; what\nExcel returns at that limit is not established here.\n\n\n", "sources": "\n- Microsoft, HSTACK function —\n \n (horizontal appending in sequence, and `#N/A` padding for arguments shorter than the tallest).\n- Handbook `content/model/01-value-universe.md` (array kind; errors as values; blank\n normalization at the published boundary).\n- Handbook `content/model/02-coercion-and-lifting.md` (error propagation at the argument\n boundary).\n- Handbook `content/model/03-call-pipeline.md` (`ValuesOnlyPreAdapter`; host-side spill\n adaptation).\n- Handbook `data/functions/FUNC.HSTACK.json` and `data/presence/FUNC.HSTACK.json` (arity,\n classification axes, implementing module).\n\n", "version-differences": "No comparison between specific Excel builds is recorded here yet. This does not establish that all builds behave alike.", "what-it-does": "\n`HSTACK(array1, [array2], …)` places its arguments side by side, left to right in the order\ngiven, and returns the combined array.\n\nThe result's **width** is the sum of the arguments' widths. Its **height** is the height of the\ntallest argument — and every argument shorter than that is padded at the bottom. Microsoft\ndocuments the pad value as `#N/A`, which is the same convention `VSTACK` uses and the same\ndefault `EXPAND` uses: synthesized cells announce themselves rather than passing as zeros.\n\nThat padding rule is what makes `HSTACK` a *stacking* operation rather than a join. It does not\nalign rows by any key; it aligns them by position, starting at row 1. Two lists of different\nlengths stacked horizontally give you a two-column array whose shorter column trails off into\n`#N/A`, which is correct and is very often not what the author meant.\n\n\n\n### Arguments\n\n\n| Argument | Meaning |\n|---|---|\n| `array1` | The first array or range. Required. |\n| `array2`, … | Further arrays, in left-to-right order. Optional, to the arity ceiling of 255 arguments. |\n\nScalars are admissible: a scalar is a 1×1 array, so `HSTACK(A1:A10, \"total\")` produces a\ntwo-column result whose second column is one value and nine `#N/A`s.\n\nThe misunderstanding to head off: **`HSTACK` takes several arguments, not a multi-area\nreference.** `HSTACK((A1:A5,C1:C5))` passes one union-shaped argument, which is a different\nthing from `HSTACK(A1:A5, C1:C5)`. What the union form does is not established here.\n\n\n" }, "signature": { "display": "HSTACK(array1, [array2], ...)", "parameters": [ { "description": null, "name": "array1", "optional": false, "repeats": false }, { "description": null, "name": "array2", "optional": true, "repeats": true } ], "trailing_repeats": true }, "sources": { "metadata": { "admission": { "label": "implemented", "reason": null }, "arity": { "max": 254, "min": 1 }, "category": "Lookup and reference functions", "classification": { "arg_preparation_profile": "ValuesOnlyPreAdapter", "coercion_lift_profile": "Custom", "determinism_class": "Deterministic", "error_collapse_profile": "None", "fec_dependency_profile": "None", "host_interaction_class": "None", "kernel_signature_class": "Custom", "rich_value_usage": "RichBlind", "surface_fec_dependency_profile": "RefOnly", "thread_safety_class": "SafePure", "volatility_class": "NonVolatile" }, "descriptions": { "long": null, "short": "Arrays placed side by side into one array", "short_source": "dna-calc-own" }, "docs": { "english_description": "Appends arrays horizontally and in sequence to return a larger array", "microsoft_url": "https://support.microsoft.com/en-us/office/hstack-function-98c4ab76-10fe-4b4f-8d5f-af1c125fe8c2" }, "entry_kind": "built_in_function", "localized_family_names": null, "localized_names": [ { "locale": "ar-SA", "name": "HSTACK" }, { "locale": "bg-BG", "name": "HSTACK" }, { "locale": "cs-CZ", "name": "SROVNAT.VODOROVNĚ" }, { "locale": "da-DK", "name": "VSTAK" }, { "locale": "de-DE", "name": "HSTACK" }, { "locale": "el-GR", "name": "HSTACK" }, { "locale": "en-AU", "name": "HSTACK" }, { "locale": "en-GB", "name": "HSTACK" }, { "locale": "en-US", "name": "HSTACK" }, { "locale": "es-ES", "name": "APILARH" }, { "locale": "et-EE", "name": "HSTACK" }, { "locale": "fi-FI", "name": "HPINO" }, { "locale": "fr-FR", "name": "ASSEMB.H*" }, { "locale": "he-IL", "name": "HSTACK" }, { "locale": "hr-HR", "name": "HSTACK" }, { "locale": "hu-HU", "name": "VÍZSZ.HALMOZÁS" }, { "locale": "id-ID", "name": "HSTACK" }, { "locale": "it-IT", "name": "HSTACK" }, { "locale": "ja-JP", "name": "HSTACK" }, { "locale": "ko-KR", "name": "HSTACK" }, { "locale": "lt-LT", "name": "HSTACK" }, { "locale": "lv-LV", "name": "HSTACK" }, { "locale": "nb-NO", "name": "HSTACK" }, { "locale": "nl-NL", "name": "HOR.STAPELEN" }, { "locale": "pl-PL", "name": "STOS.POZ" }, { "locale": "pt-BR", "name": "HSTACK" }, { "locale": "pt-PT", "name": "JUNTARH" }, { "locale": "ro-RO", "name": "HSTACK" }, { "locale": "ru-RU", "name": "ГСТОЛБИК" }, { "locale": "sk-SK", "name": "HSTACK" }, { "locale": "sl-SI", "name": "HSTACK" }, { "locale": "sr-Latn-RS", "name": "HSTACK" }, { "locale": "sv-SE", "name": "HSTACK" }, { "locale": "th-TH", "name": "HSTACK" }, { "locale": "tr-TR", "name": "YATAYYIĞ" }, { "locale": "uk-UA", "name": "HSTACK" }, { "locale": "vi-VN", "name": "HSTACK" }, { "locale": "zh-CN", "name": "HSTACK" }, { "locale": "zh-HK", "name": "HSTACK" }, { "locale": "zh-TW", "name": "HSTACK" } ], "signature": { "display": "HSTACK(array1, [array2], ...)", "parameters": [ { "description": null, "name": "array1", "optional": false, "repeats": false }, { "description": null, "name": "array2", "optional": true, "repeats": true } ], "trailing_repeats": true }, "surface_name": "HSTACK", "version_marker": null, "xlcall": null }, "microsoft": "https://support.microsoft.com/en-us/office/hstack-function-98c4ab76-10fe-4b4f-8d5f-af1c125fe8c2" }, "try_it": { "formula": "=HSTACK({1;2},{3;4})" }, "url": "https://handbook.dnacalc.com/functions/HSTACK.html", "variants": [], "versions": [ { "behaviour": "=HSTACK(1,#VALUE!) returned #VALUE!. =HSTACK(Data!C1,{1;2}) returned {#N/A,1;#N/A,2}. Inputs were supplied through cells using Value2, formula literals. These are controls from this environment, not a same-input comparison with another build.", "excel_build": "16.0.20527.20088", "observations": [ { "call": "=HSTACK(1,#VALUE!)", "excel": { "bits": null, "display": "#VALUE!", "kind": "error" }, "input_route": "formula-literal", "inputs": [], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA" }, { "call": "=HSTACK(Data!C1,{1;2})", "excel": { "bits": null, "display": "{#N/A,1;#N/A,2}", "kind": "array" }, "input_route": "cell-value2", "inputs": [ { "name": "Data!C1", "value": { "bits": null, "display": "#N/A", "kind": "error" } } ], "recorded_in": "Excel 16.0.20527.20088 Current Channel, Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA" } ], "platform": "Windows 64-bit (NT 10.0.26200), x86-64, AMD Ryzen 7 4800H, en-ZA", "sources": [ "https://github.com/DnaCalc/ExcelFunctionsHandbook/blob/3345a87/content/notes/FN-HSTACK-001.json" ] } ] }