The same trigonometric keys can encode two different questions
The expression sin(30) is incomplete unless its angle unit is known. In degree mode it represents the sine of 30°, which is 0.5. In radian mode it represents the sine of 30 radians, a very different point on the unit circle. OpenStax explicitly advises setting the intended mode before evaluating trigonometric functions. The Deg or Rad label on this workbench is therefore part of the input, not a display preference.
Inverse trigonometric results depend on the same setting. If a calculation uses measurements from a drawing, physics formula, spreadsheet, or code library, copy the source's angle convention into your notes. Converting an angle can be valid; silently switching the calculator mode midway through a chain is not.
Function names carry domain rules
An error is often useful information that the entered value lies outside the supported real-number question.
Logarithms require a valid base and argument
For real logarithms the argument must be positive, while a custom base must be positive and cannot equal one. A domain error should prompt a review of the model, not a guessed replacement value.
Roots depend on the index
An odd root can return a real result for a negative radicand; an even root cannot. Writing the index and radicand separately prevents a minus sign from being applied at the wrong level.
Factorials are discrete here
The factorial key accepts supported non-negative whole numbers. It does not evaluate the gamma function or extend factorial notation to arbitrary real or complex inputs.
Tangent has singular directions
Near an odd multiple of 90° or π/2, a finite decimal approximation can become enormous. Check the exact angle and mode instead of reading that magnitude as a stable physical result.
Enter nested expressions as a structure, not a keystroke story
A reproducible entry preserves which operation owns each argument.
- Group the innermost quantity
For ln(3 + √5), finish 3 + √5 inside the logarithm before closing the outer function. The visible parentheses should match the structure written in the source formula.
- Separate exponent signs from subtraction
Write 4.7 × 10⁻⁶ with the EE control or explicit grouped exponent. A negative exponent scales one number; it does not subtract six from the preceding expression.
- Retain guard digits until the end
Carry the calculator's unrounded intermediate value through later operations and round only the reported output. Re-entering a short display at every stage can accumulate avoidable error.
Grouped arithmetic keeps a technical expression auditable
The ordinary arithmetic keys use the same expression tree as the scientific functions, so grouping and operator order remain part of the input.
Trace the groups before the outer operation
For (18.6−4.2)² ÷ (3×1.5+0.3), resolve the grouped difference to 14.4 and the grouped denominator to 4.8. Squaring gives 207.36, and the final division gives 43.2. Removing either pair of parentheses asks a different question.
Make slash scope and negative bases explicit
Plain-text /(3*1.5+0.3) preserves the complete denominator. Likewise, −4² means the negative of 4² under the usual convention, while (−4)² squares a negative base. Parentheses are more reliable than assuming every reader or tool will infer the intended scope.
Treat percent as an operand, not a hidden business rule
The percent key divides the active value by 100. Thus 200×5% is 10, while a five-percent increase must be entered as 200×(1+5%) to produce 210. The calculator does not silently reinterpret addition as a markup operation.
Choose a specialist when evaluation becomes analysis
- Use the Exponent Calculator when the main task is comparing power forms, negative exponents, roots, or step-by-step exponent rules rather than embedding one power in a long expression.
- Open the Logarithm Calculator to solve a focused log value with the base, domain checks, change-of-base relationship, and interpretation kept together.
- Move to the Limit Calculator when the question concerns behavior as an input approaches a point or grows without bound. Substituting one very large finite input is not a proof of a limit.
- For ordinary shopping, totals, memory, and a compact history, the Basic Calculator keeps the interface narrower and the daily arithmetic easier to inspect.
Approximation, overflow, and symbolic equality are separate claims
Constants such as π and e are represented by finite browser numbers, and irrational results such as √2 are displayed as rounded approximations. A screen showing 1.41421356237 is useful numeric evidence, but it is not the exact object √2. Likewise, a huge power can exceed the Number range even when the mathematical integer itself is finite.
The workbench rejects undefined and non-finite ordinary results. It does not solve equations, simplify symbols, track units, estimate measurement uncertainty, or certify a scientific model. Save the expression, mode, source values, and final rounding rule whenever another person must reproduce the result.