Roman Numeral Converter

Definition

Roman numerals assign integer values to seven Latin letters: I=1, V=5, X=10, L=50, C=100, D=500, M=1000. A numeral is a finite string of those marks read under additive and subtractive rules. In schoolbook (canonical) convention the only subtractive pairs are IV, IX, XL, XC, CD, and CM; otherwise a smaller mark does not precede a larger one. The toRoman function on this page is a bijection from {1,…,3999} onto the set of canonical strings from I to MMMCMXCIX.

The inverse first sums tokens greedily from the left (longest matching token from the MAP list), then requires equality with toRoman(n). That second check rejects non-canonical forms even when a naive letter sum would yield a “correct” arithmetic value (for example IIII = 4 arithmetically, but canonically 4 = IV).

This calculator’s domain excludes zero, fractions, negatives, and integers ≥ 4000. Classical Roman notation had no zero symbol; larger values were historically marked with a vinculum (overline, ×1000) among other extensions — outside this calculator’s range.

Historical and conceptual context

Roman notation grows out of earlier Italic and Etruscan practices; marks resembling I, V, and X appear in epigraphy before the alphabet was fully Latinized. Karl Menninger in Number Words and Number Symbols traces the path from tally strokes and “fives” shaped like a hand or angle to the seven-letter set used in Roman administration and inscriptions. Georges Ifrah in The Universal History of Numbers stresses that the system was a recording and public-accounting tool, not a positional base-ten system — hence the absence of zero and a different logic from Hindu–Arabic digits.

Stephen Chrisomalis in Numerical Notation: A Comparative History places Roman numerals among cumulative–additive systems with later subtractive rules. Subtractive spelling was not uniform from the start: inscriptions and medieval manuscripts show additive forms (IIII, VIIII) beside subtractive ones (IV, IX). The modern schoolbook set of six subtractive pairs is a didactic and typographic convention solidified in print culture and schooling — and that is the convention this page implements.

Tower and dial clocks often keep IIII instead of IV: explanations range from visual balance against VIII, to workshop tradition, to avoiding a form associated with Jupiter (IVPITER) in some accounts — whatever the cause, clockmaker IIII is not canonical here. In film, monuments, and papal names (John Paul II as Ioannes Paulus II) Roman numerals today serve symbolic and stylistic roles rather than everyday arithmetic.

Formal properties

Let T be the ordered list of (value, token) pairs: (1000,M), (900,CM), (500,D), (400,CD), (100,C), (90,XC), (50,L), (40,XL), (10,X), (9,IX), (5,V), (4,IV), (1,I). For n ∈ {1,…,3999} the greedy algorithm builds a unique string R(n) by appending tokens until the summed value equals n. The inverse: scanning string s with tokens from T left to right yields sum σ(s); acceptance requires σ(s) ∈ [1,3999] and R(σ(s)) = s (after uppercase and space stripping).

Properties: R is injective and surjective onto its image; |R(n)| is bounded (the longest canonical strings in this range are on the order of a dozen characters, e.g. MMMCMXCIX). Tokens V, L, and D never repeat in R(n); I, X, C, and M repeat at most three times in a row in additive stretches. Each subtractive pair appears at most once per decimal “column.”

The system is not positional: a mark’s meaning does not depend on a base power by place, but on local neighborhood rules (add / subtract). There is therefore no direct analogue of a decimal place shift.

Full derivation

Step 1 (Arabic → Roman). Input: integer n with 1 ≤ n ≤ 3999. Remainder r ← n, output w ← empty. For each (v, t) in T: while r ≥ v, append t to w and set r ← r − v. Output: w. Example n = 2026: M, M, X, X, V, I → MMXXVI.

Step 2 (Roman → Arabic). Input: string s. Normalize: uppercase, strip spaces. Reject if any character is outside MDCLXVI. Index i ← 0, sum σ ← 0. For each (v, t) in T: while s matches t at i, set σ ← σ + v and i ← i + |t|. If i ≠ |s|, reject. If σ ∉ [1,3999] or R(σ) ≠ s, reject. Otherwise the result is σ.

Step 3 (why canonical validation). A naive sum of IIII would give 4, but R(4) = IV ≠ IIII. A naive IC would give 99, but R(99) = XCIX ≠ IC. Requiring R(σ) = s accepts only strings generated by the same model.

Step 4 (subtractive example). n = 44: XL (40) + IV (4) = XLIV. Tokens XL and IV come from T; XXXXIIII is not produced.

Special cases

Lower end: 1 → I. Upper end: 3999 → MMMCMXCIX. Round values: 10 → X, 50 → L, 100 → C, 500 → D, 1000 → M, 2000 → MM. Fours and nines in each decade: 4→IV, 9→IX, 40→XL, 90→XC, 400→CD, 900→CM. Year 1990 → MCMXC; 2026 → MMXXVI; 14 → XIV.

Rejected: 0, 4000, −3, 14.5, empty string, “ABC”, “IIII”, “VV”, “IC”, “IL”, “XM”. Lowercase “xiv” is accepted as XIV. Spaces inside “X I V” are stripped before reading.

Vinculum (M with an overline = 1,000,000 in some conventions) and apostrophe/parenthesis thousands notations are unsupported — values beyond 3999 report a range error.

Numerical remarks

The calculator is integer-only: no fractions appear at any step. Arabic input after comma-to-dot normalization must be a finite whole number (n === ⌊n⌋). Roman reading does not “blind-sum” without the canonical check — a deliberate choice matching schoolbook spelling.

Complexity: with a fixed-length list T and a short string (a dozen characters), runtime is effectively constant. There is no int32 overflow risk in 1–3999.

The result panel shows direction, input echo, subtractive pairs present in the string, and greedy-breakdown chips — the same values as in the algorithm, without a separate visual heuristic.

Related notions and sources

Related CalcBoxer tools: positional bin/hex/dec (IT), random number generator, decimal-to-fraction. Roman numerals are neither a cipher nor a date calculator — Roman-style calendar dates (day.month.year as three separate numbers) need a separate mode this calculator does not provide.

Sources Menninger, Number Words and Number Symbols; Ifrah, The Universal History of Numbers; Chrisomalis, Numerical Notation; encyclopedic entries Wikipedia EN/PL “Roman numerals” / “Cyfry rzymskie”; school tables of the IV–CM pairs. Educational tool — not a substitute for epigraphic paleography.