Mathematics & Conversions

Data Normalization Calculator — Min-Max, Z-Score & Robust Scaling

Scale a single numeric value using Min–Max normalization, Z-score standardization, or robust median/IQR scaling with user-supplied reference statistics.

Unit-aware inputs Deterministic calculation Engineering interpretation
Calculation workspace

Enter the known values and review the calculated result

Deterministic calculation
01
Parameters

Input parameters

Use consistent values and select the intended engineering units.

Scaling method

Value to scale

Min–Max reference parameters

Z-Score reference parameters

Robust scaling reference parameters

02
Output

Results

Live
Ready to calculate Complete the required inputs and run the calculation.
—
Engineering Pro Save, document and continue this calculation

Turn this result into a reusable engineering record with saving, PDF export and reporting workflows.

Engineering reference

Method, application and limitations

Review the calculation method, intended application and engineering assumptions before using the result in a design decision.

01
Method

Formula and calculation method

This calculator scales a single input value x using Min–Max normalization, Z-score standardization, or robust scaling. The reference statistics are entered directly; the calculator does not derive them from a dataset.

Min–Max normalization formula

Δxsource = xmax − xmin

Δxtarget = tmax − tmin

r = (x − xmin) / (xmax − xmin)

k = (tmax − tmin) / (xmax − xmin)

x* = tmin + r · (tmax − tmin)

Equivalently:

x* = tmin + (x − xmin) · k

If the target limits are left blank, the calculator uses tmin = 0 and tmax = 1.

Z-score standardization formula

Δx = x − μ

k = 1 / σ

x* = z = (x − μ) / σ

The result is the standard score of the entered value relative to the supplied mean and standard deviation.

Robust scaling formula

Δx = x − x̃

k = 1 / IQR

x* = (x − x̃) / IQR

Variables

  • x* — scaled value reported by the calculator
  • x — input value
  • xmin — reference minimum for Min–Max scaling
  • xmax — reference maximum for Min–Max scaling
  • tmin — target minimum; default 0 when omitted
  • tmax — target maximum; default 1 when omitted
  • r — relative position of x within the reference Min–Max range
  • μ — reference mean for Z-score standardization
  • σ — reference standard deviation; must be greater than zero
  • x̃ — reference median for robust scaling
  • IQR — reference interquartile range; must be greater than zero
  • k — scaling coefficient used by the selected transformation
  • Δx — offset from the applicable reference value

The calculator performs numerical scaling only and does not perform unit conversion. Values that describe the same physical quantity must therefore use consistent units. The primary scaled result is reported as dimensionless.

02
Application

When to use this calculator

Use this calculator when you already know the reference parameters for a value and want to transform that value to a different numerical scale.

  • Use Min–Max normalization to map a value from a known reference interval into a target interval such as 0 to 1 or another user-defined range.
  • Use Z-score standardization when the reference mean and standard deviation are known and you want to express the value as its signed number of standard deviations from the mean.
  • Use robust scaling when the reference median and interquartile range are available and the required transformation is based on median/IQR centering and scaling.
  • Use Min–Max mode to check whether a new value lies inside or outside the supplied reference minimum and maximum. Values outside that interval are still calculated but are marked as extrapolation.
  • Use a custom Min–Max target range when a value must be linearly mapped to limits other than the default 0 to 1 range.

This is a single-value scaling calculator. It does not accept a complete dataset and does not calculate the minimum, maximum, mean, standard deviation, median, quartiles, or IQR from raw observations. These reference values must be obtained separately and entered into the appropriate fields.

The calculator also does not calculate Z-score percentiles or probabilities, classify statistical outliers, clip Min–Max results to the target range, or assess whether a dataset follows a particular probability distribution.

03
Decision support

How to interpret the result

x* is the transformed value produced by the selected scaling method. Its interpretation depends on whether Min–Max, Z-score, or robust scaling is selected.

Min–Max result

In Min–Max mode, the reference minimum maps to the target minimum and the reference maximum maps to the target maximum. A value between the two reference limits maps proportionally inside the target interval.

  • r = 0 corresponds to x = xmin.
  • r = 1 corresponds to x = xmax.
  • 0 < r < 1 means the value lies inside the reference range.
  • r < 0 or r > 1 means the value lies outside the reference range and the calculation is an extrapolation.

The calculator does not clip an extrapolated result. Therefore, with the default target range of 0 to 1, an input below xmin can produce x* < 0 and an input above xmax can produce x* > 1.

Z-score result

In Z-score mode, x* is the standard score z = (x − μ) / σ. A result of 0 means the entered value equals the supplied mean. Positive values are above the mean and negative values are below the mean. For example, z = 1.5 means the input is 1.5 supplied standard deviations above the supplied mean.

The calculator does not convert the Z-score to a percentile, probability, p-value, or automatic outlier classification.

Robust scaling result

In robust mode, x* = (x − x̃) / IQR. A result of 0 means the input equals the supplied median. A result of 1 means the input is one supplied IQR above the median, while −1 means it is one supplied IQR below the median.

Robust mode performs the median/IQR transformation only. The current calculation does not apply an automatic outlier threshold to the resulting value.

Calculator status

  • Safe — the required inputs are valid and the calculation is finite. In Min–Max mode, the input is also inside the entered reference range. In Z-score and robust modes, this status indicates a valid calculation rather than a statistical acceptance criterion.
  • Warning — generated in Min–Max mode when x is below xmin or above xmax. The result is still calculated, but it represents extrapolation beyond the declared reference range.
  • Invalid — generated when required values are missing or non-finite, xmax ≤ xmin, tmax ≤ tmin, σ ≤ 0, IQR ≤ 0, or a calculated primary/intermediate result is not finite.

A valid or safe calculator status confirms that the selected mathematical transformation was evaluated successfully. It does not by itself establish data quality, statistical significance, model suitability, or engineering safety.

04
Worked case

Calculation example

Example 1 — Min–Max normalization to 0–1

A sensor value of 75 is to be scaled using a reference range from 50 to 100. The default target range of 0 to 1 is used.

  • x = 75
  • xmin = 50
  • xmax = 100
  • tmin = 0
  • tmax = 1

Δxsource = 100 − 50 = 50

Δxtarget = 1 − 0 = 1

r = (75 − 50) / 50 = 0.5

k = 1 / 50 = 0.02

x* = 0 + 0.5 · 1 = 0.5

The input lies halfway between the reference minimum and maximum, so it maps to 0.5 in the 0–1 target range. Because 75 is inside the reference interval, no extrapolation warning is generated.

Example 2 — Z-score standardization

An input value of 72 is evaluated using a reference mean of 60 and a standard deviation of 8.

  • x = 72
  • μ = 60
  • σ = 8

Δx = 72 − 60 = 12

k = 1 / 8 = 0.125

x* = z = 12 / 8 = 1.5

The entered value is therefore 1.5 supplied standard deviations above the supplied mean. The calculator does not convert this result to a percentile or probability.

Example 3 — Robust scaling with median and IQR

An input value of 68 is scaled using a reference median of 60 and an interquartile range of 16.

  • x = 68
  • x̃ = 60
  • IQR = 16

Δx = 68 − 60 = 8

k = 1 / 16 = 0.0625

x* = 8 / 16 = 0.5

The input is 0.5 IQR above the supplied median on the calculator’s robust scaling scale. This result is a transformation value, not an automatic outlier classification.

05
Model boundaries

Assumptions and limitations

  • The calculator transforms one finite numeric input value at a time.
  • Reference statistics are supplied by the user. The calculator does not calculate minimum, maximum, mean, standard deviation, median, quartiles, or IQR from a dataset.
  • The calculator performs no unit conversion. Values representing the same physical quantity must use consistent units.
  • Min–Max scaling requires xmax > xmin.
  • Min–Max target limits must satisfy tmax > tmin.
  • If Min–Max target limits are omitted, the calculator uses a target minimum of 0 and a target maximum of 1.
  • Min–Max results are not clipped to the target range. Inputs outside the reference range are extrapolated and marked with a warning.
  • Z-score standardization requires a finite mean and a finite standard deviation greater than zero.
  • Robust scaling requires a finite median and a finite IQR greater than zero.
  • Z-score mode performs only the transformation (x − μ) / σ. It does not calculate percentiles, probabilities, p-values, or statistical significance.
  • Robust mode performs only the transformation (x − x̃) / IQR. It does not automatically classify an observation as an outlier.
  • The current warning logic applies only to Min–Max extrapolation. No magnitude-based warning threshold is applied to Z-score or robust results.
  • The validity status confirms numerical consistency with the implemented transformation; it does not validate the origin, representativeness, or statistical suitability of the supplied reference parameters.
06
Questions

Frequently asked questions

What is the Min–Max normalization formula?
Min–Max normalization is calculated as x* = t_min + [(x − x_min) / (x_max − x_min)] · (t_max − t_min), where x is the value to scale, x_min and x_max define the reference range, and t_min and t_max define the target range. If the target limits are omitted, this calculator uses 0 and 1.

How do you normalize a value between 0 and 1?
For a 0–1 target range, use x* = (x − x_min) / (x_max − x_min). A value equal to x_min gives 0, a value equal to x_max gives 1, and values between the limits produce results between 0 and 1. This requires x_max to be greater than x_min.

What is the difference between Min–Max normalization and Z-score standardization?
Min–Max normalization maps a value relative to a reference minimum and maximum and can place values inside a selected target interval. Z-score standardization instead subtracts the reference mean and divides by the reference standard deviation, producing a standard score that expresses the value’s position relative to the mean in units of the supplied standard deviation.

How is a Z-score calculated?
The Z-score is calculated as z = (x − μ) / σ, where x is the input value, μ is the reference mean, and σ is the reference standard deviation. A Z-score of 0 means x equals the mean, a positive score is above the mean, and a negative score is below the mean. This calculator requires σ to be greater than zero.

Does the Z-score mode calculate a percentile or probability?
No. The calculator returns only the standardized value z = (x − μ) / σ and its numerical scaling intermediates. It does not apply a normal-distribution CDF and does not calculate percentiles, tail probabilities, p-values, or statistical significance.

What is robust scaling and how is it calculated?
In this calculator, robust scaling centers the input on the supplied median and divides the result by the supplied interquartile range: x* = (x − x̃) / IQR. A result of 0 corresponds to the median, while positive and negative values indicate positions above or below the median on the entered IQR scale.

Can a Min–Max normalized value be less than 0 or greater than 1?
Yes. With a 0–1 target range, values remain between 0 and 1 only when x is between the entered reference minimum and maximum. If x is below x_min or above x_max, this calculator extrapolates the linear transformation instead of clipping it, so the result can be below 0 or above 1. The calculator marks this condition with a warning.

Can I scale to a range other than 0 to 1?
Yes. In Min–Max mode, enter a custom target minimum and target maximum. The calculator linearly maps x_min to the target minimum and x_max to the target maximum. The target maximum must be greater than the target minimum.

Does this calculator normalize an entire dataset?
No. This calculator transforms one input value using reference parameters that you provide. It does not read a dataset or calculate its minimum, maximum, mean, standard deviation, median, quartiles, or IQR. Calculate those reference statistics separately before using this tool.

Does robust scaling automatically detect outliers?
No. Robust mode calculates (x − median) / IQR but does not apply an automatic outlier threshold. The returned scaled value should not be interpreted as an outlier classification unless a separate statistical criterion is applied.

When is the normalization or scaling calculation invalid?
The calculation is invalid when required values are missing or non-finite. Min–Max mode also requires x_max > x_min and t_max > t_min. Z-score mode requires σ > 0, and robust scaling requires IQR > 0. The calculator also rejects a result if the scaled value or required intermediate values are not finite.

Add an engineering note
Was this calculator useful?
Engineering notice. Verify input data, assumptions, applicable standards and required safety factors before using results for design release or safety-critical decisions.