Pro Calc: Exact Fractions, Units, Matrices and Scripting

Pro Calc is a browser expression calculator: type a formula and it is parsed with mathematical precedence — 2 + 3 4 reads 14, while (2 + 3) 4 reads 20. It accepts what a keypad calculator cannot: named variables and functions, units with conversion, complex numbers, matrices and a 31-entry scientific-constants panel. Three mode buttons choose the arithmetic engine behind the same input line — Standard (binary64), Fraction (exact rationals) and BigNumber (64 significant digits) — and each mode keeps its own variables and Ans. A Script JS tab runs sandboxed JavaScript with those libraries exposed by name. Everything computes in the browser.

Updated 2026-10-01Reviewed by CalcX Editorial Team5 sources5 min read

What Pro Calc is

Pro Calc is a browser expression calculator: type a formula and it is parsed with mathematical precedence — 2 + 3 * 4 reads 14, while (2 + 3) * 4 reads 20. It accepts what a keypad calculator cannot: named variables and functions, units with conversion, complex numbers, matrices and a 31-entry scientific-constants panel. Three mode buttons choose the arithmetic engine behind the same input line — Standard (binary64), Fraction (exact rationals) and BigNumber (64 significant digits) — and each mode keeps its own variables and Ans. A Script JS tab runs sandboxed JavaScript with those libraries exposed by name. Everything computes in the browser.

How to use it

Type an expressionon the input line — 1/3 + 1/6, sqrt(-4), det([1,2;3,4]) or a unit sum such as 5 cm + 2 inch to mm. The keypad inserts the same tokens, Ans recalls the last result, and to converts units.
Choose a number mode first.Standard, Fraction (exact) and BigNumber (64 digits) are separate engines; variables and Ans belong to the mode that created them.
Name values and functions.a = 3^2 returns 9; f(x) = x^2 + 1 reports Function defined; f(Ans) then returns 82. The Variables panel lists every name and can delete one.
Pull in a constantfrom the 31-entry panel — pi, speedOfLight or gravity — which inserts the name; the panel previews values to 8 significant digits.
Open Script JSfor batched work: math, cas, stats, fin, logic, graph, latex, currency, date and family, with print() for output lines and return for the final value.

Worked examples

Every readout below comes from this page; type the same input in the same mode.

Mode	Input	Readout
Standard	`1/3 + 1/6`	0.5
Fraction	`1/3 + 1/6`	1/2
BigNumber	`1/3 + 1/6`	0.5
Standard	`0.1 + 0.2 - 0.3`	5.5511151231258e-17
Standard	`5 cm + 2 inch to mm`	100.8 mm
Standard	`9.81 m/s^2 * 5 s`	49.05 m / s
Standard	`(3+4i) * (1-2i)`	11 - 2i
Standard	`det([1,2;3,4])`	-2
Fraction	`1.5 + 1/3`	11/6
Standard	`pi`	3.1415926535898
BigNumber	`pi`	3.141592653589793238462643383279502884197169399375105820974944592
Standard, ENG on	`12345`	12.345000000000e+3

Two details behind the table: unit sums carry their unit (5 cm + 2 inch reads 10.08 cm before conversion), and the BigNumber pi row shows 64 significant digits of the stored value — the next two sections cover what that does and does not mean.

Limits to keep in mind

Display rounding.: Standard mode prints 14 significant digits; 0.1 + 0.2 reads 0.3, yet 0.1 + 0.2 - 0.3 exposes the stored binary64 value, 5.5511151231258e-17. ENG pads the display to 14 significant digits (12.345000000000e+3), and the constants panel previews only 8.
Variables and `Ans` are mode-local.: r = 3^2 in Standard reads 9 there; switch to Fraction and typing r stops with Undefined symbol r. The history list is shared across modes, but switching modes never converts values or moves names.
Fraction mode is for rational arithmetic.: sqrt(4) stops with a Fraction/BigNumber conversion message, while the rational power 4^(1/2) returns the exact 2/1; use Standard or BigNumber for root functions.
What it does not do: it is not a plotting or step-by-step tool.: The keypad view has no graph and no derivation trace; plot with the graphing calculator, and use Math Tools for symbolic derivatives, integrals and rule-by-rule steps. For plain four-function arithmetic, the standard calculator is the shorter path.

The mode is an engine, not a display filter

The signature trap is that the three mode buttons are three different calculations under one input line. 1/3 * 3 reads 1 in Standard, 1/1 in Fraction — and 0.9999999999999999999999999999999999999999999999999999999999999999 (64 nines) in BigNumber, because BigNumber rounds 1/3 to 64 significant digits before multiplying by 3. The extra digits are honest arithmetic on a rounded input, not a better answer. Pick Fraction when the exact ratio is the point, Standard when binary64 behavior is acceptable, and BigNumber when the inputs themselves are long decimals.

Where it fits

Checking rational arithmetic

Fraction mode keeps intermediates as ratios, so an answer can be compared exactly: 1/3 + 1/6 is 1/2, 7/9 * 3/7 reduces to 1/3, 2/4 to 1/2 and 2^10 to 1024/1.

Unit-aware engineering arithmetic

Units evaluate as quantities: 9.81 m/s^2 * 5 s is 49.05 m / s and 2 hours + 45 minutes to minutes is 165 minutes. The inch is the international inch, exactly 25.4 mm, which is why 5 cm + 2 inch to mm lands on 100.8 mm — no rounding tail.

Batch work and cross-checking in the script tab

The built-in cross-library example factors x² − 5x + 6 to (-2+x)*(-3+x) and solves it as [2,3], fits [[1,2.1],[2,4.2],[3,8.9],[4,16.2]] to y = 1.0693·e^(0.68802x) with R² 0.9953629406545152, and returns the antiderivative -2*sin(x)+2*cos(x)*x+sin(x)*x^2 + C. A finance example scans a 30-year loan of 1,000,000 at 3.0%–6.0% and prints monthly payments from 4216.04 to 5995.51; a cross-check example compares a symbolic integral (pi) with a numeric one (3.1415926535914935) and returns MATCH ✓.

References

mathjs documentation, Expressions — syntax, mathjs.org （访问日期：2026-10-01）— precedence and expression syntax.
mathjs documentation, Units, mathjs.org （访问日期：2026-10-01）— unit arithmetic and to conversion.
mathjs documentation, Fractions, mathjs.org （访问日期：2026-10-01）— exact rational arithmetic in Fraction mode.
NIST, Special Publication 811: Guide for the Use of the International System of Units (SI), nist.gov （访问日期：2026-10-01）— the international inch equals exactly 25.4 mm.
IEEE, IEEE 754-2019 — IEEE Standard for Floating-Point Arithmetic, standards.ieee.org （访问日期：2026-10-01）— binary64 storage and rounding in Standard mode.