# Domain and Range: Definition, Examples and Uses in Lab Data

By the end of this guide you will be able to state the domain and range of a function in interval notation, read both from a graph, and apply the same reasoning to a [qPCR standard curve](/blog/guides/qpcr-standard-curves-how-to-assess-efficiency-and-quantification-limits), a [dose response curve](/blog/research-skills/ic50-vs-ec50-dose-response-curve), a pH reading, and a spectrophotometer reading. You will also be able to spot the moment a fit, a dilution, or a detector has pushed your data outside the region where the numbers mean anything. Have on hand a spreadsheet or a Python environment with NumPy and pandas, your current calibration data, and a [domain and range calculator](/blog/data-analysis/calculator-functions-for-molecular-biology-lab) if you want a fast check on the algebra.

## The Domain and Range Definition in Plain Language

<!-- art-pending: scripts/editorial-illustration/specs/domain-and-range-benchtop-spectrophotometer.json -->

A function is a rule that assigns exactly one output to each input it accepts. The domain is the set of inputs the rule accepts. The range is the set of outputs the rule actually produces.

Mathematicians write a function as f: X to Y. Here X is the domain, Y is the codomain, and the range is the subset of Y that f reaches. The codomain is the declared output space. The range is what is realized. A function that converts Celsius to Fahrenheit has a codomain of all real numbers and, over any real sample, a range that is also all real numbers. A function that returns a probability has a codomain of [0, 1] and a range that sits inside it.

The vocabulary extends to families of functions. When a collection of functions shares an input set and an output set, those sets are called the common domain and common range [1]. A lab assay has the same structure. The set of concentrations the method can accept is the domain. The set of signals the detector can report is the range.

Two small examples fix the idea.

f(x) = 1/x. The domain is every real number except 0, written (-inf, 0) U (0, inf). The range is the same set. No nonzero input produces 0, and every nonzero output is produced by its reciprocal.

f(x) = x squared. The domain is (-inf, inf). The range is [0, inf). Negative inputs are accepted and produce non-negative outputs.

### Domain, Codomain, and Range Are Three Different Sets

Keep them separate in a report. The domain is the input restriction. The codomain is the declared output type. The range is the output set that was observed or computed. Most lab write-up errors come from treating the codomain as the range, which overstates what the data can show.

### Why Lab Functions Are Almost Always Partial

A function that is not defined everywhere is called a partial function. Almost every measurement method produces one. The limit of detection removes low inputs from the domain. The upper saturation limit removes high inputs. The usable linear interval removes both ends at once. A method that cannot accept a sample is telling you the sample is outside the domain, not that the answer is zero.

## Domain and Range Notation

Interval notation is the default in lab reports and in most software output. A square bracket includes the endpoint. A parenthesis excludes it. Infinity always takes a parenthesis.

| Notation | Example | Meaning |
| Interval | (0, 1] | greater than 0, less than or equal to 1 |
| Interval | [0, inf) | greater than or equal to 0 |
| Set builder | {x in R such that x > 0} | all positive reals |
| Inequality | x > 0 | same set |
| Function with restriction | f(x) = log10(x), x > 0 | domain stated beside the rule |
| Union | (-inf, 0) U (0, inf) | two disjoint pieces |

Use the union symbol to join pieces. The domain of 1/x is (-inf, 0) U (0, inf). The range of 1/(x squared - 4) is (-inf, -0.25] U (0, inf), which mixes a closed endpoint with open ones.

The same conventions carry into code and figure captions. A caption that reads "[standard curve](/knowledge/diagnostics/molecular/interpret-bca-assay-standard-curve), N > 0" is a domain statement. A methods sentence that reads "absorbance was kept between 0.1 and 1.0 AU" is a range statement about the instrument.

### Notation Warning

Some calculators print `R` or `Reals` for (-inf, inf). Some print the domain as a chain of inequalities. Convert to interval notation before pasting into a report, because mixed formats invite misreading. A domain written as `x > 0, x ≠ 5` is clearer as (0, 5) U (5, inf).

## Reading a Domain and Range Graph

To read a domain and range graph, project the curve onto each axis. The horizontal shadow is the domain. The vertical shadow is the range.

Rules that cover most lab plots:

- A solid dot includes the point. An open dot excludes it.
- A vertical asymptote is a hole in the domain.
- A horizontal asymptote is a value the range approaches but may not reach.
- A single missing point is a removable discontinuity, and it is still a domain hole.
- A plateau at the top of a sigmoid curve is the range saturating.

Example. The curve y = 1/(x - 2) + 3 has a vertical asymptote at x = 2 and a horizontal asymptote at y = 3. Domain: (-inf, 2) U (2, inf). Range: (-inf, 3) U (3, inf).

Example. The curve y = log10(x) has a vertical asymptote at x = 0. Domain: (0, inf). Range: (-inf, inf). Every horizontal line crosses the curve exactly once.

A calibration curve is read the same way. The horizontal axis is the concentration domain of the standards. The vertical axis is the signal range of the detector. The fitted line is only drawn across the domain that was actually measured.

Bounded ranges appear wherever a score is normalized. A similarity or quality function can be designed so its range is the unit interval [0, 1], which keeps comparisons interpretable across samples even when the underlying variables have no natural algebraic or topological structure [2]. That normalization is a design decision, not a mathematical necessity, and it is worth stating explicitly in a methods section.

## Domain and Range of a Function: Four Lab Cases

### Case 1. The Domain of a Log Function in qPCR

The standard curve is

Cq = m times log10(N) + b

where N is the starting copy number, m is the slope, and b is the intercept at one copy. A slope near -3.32 corresponds to an amplification efficiency near 100 percent. The log10 is defined only for N > 0. Zero copies means no template, which cannot be placed on a log axis. A negative copy number has no physical meaning.

Domain: (0, inf) in copies per reaction.
Range of the model: (-inf, inf). Range of the instrument: finite. Most real time cyclers stop at 40 cycles, so a reported Cq sits inside a practical window.

The domain consequence is practical. A well that returns "undetermined" has no Cq, so it lies outside the domain of the fitted function. Software either drops the well or substitutes the limit of detection, and those two choices produce different reported concentrations. Decide which rule your lab follows before the run, not after.

### Case 2. Dose Response Curve Domain

The four parameter logistic model is

y = bottom + (top - bottom) / (1 + (x / EC50) raised to the hill power)

Domain: x > 0 when the x axis is a log concentration, which is the normal presentation for a dose response curve. The model also requires x > 0 when the hill slope is not an integer, because a fractional power of a negative number is not defined in the reals.

Range: for a positive hill slope the response rises from bottom to top, so the range is the open interval (bottom, top) at finite x. For a negative hill slope the curve falls and the range is (top, bottom). The asymptotes are approached, not reached.

A zero dose belongs to a separate control group. It cannot sit on the fitted curve, and most graphing software will either refuse to place it on a log axis or silently drop it.

### Case 3. pH

pH = negative log10 of hydrogen ion activity.

Domain: hydrogen ion activity greater than zero.
Range: (-inf, inf). Negative pH values exist in concentrated strong acids, and values above 14 exist in concentrated bases. The familiar 0 to 14 window is the routine aqueous working range, not the mathematical range.

Practical point. Report the calibration range with the reading. A pH value of 13.7 from an electrode calibrated at 4.00, 7.00 and 10.00 is an extrapolation, and the uncertainty is larger than the display suggests.

### Case 4. Absorbance and Beer-Lambert

![Worked steps taking a 1.85 AU sample through a factor of four dilution to 0.46 AU and back to 1.84 AU.](/article-images/illustrations/domain-and-range-absorbance-dilution-calc.webp)

A = epsilon times c times l, where epsilon is the molar absorptivity, c is concentration, and l is path length.

The equation is linear over a finite window. Below roughly 0.1 AU the difference between sample and blank is small relative to detector noise, so the signal to noise ratio is poor. Above roughly 1.0 AU stray light and detector nonlinearity compress the signal, so the computed concentration is biased low. The domain of the calibration is the concentration span of the standards. The valid range of [absorbance readings](/knowledge/diagnostics/molecular/calculate-dna-concentration-absorbance) is the linear window of the instrument.

A sample that reads outside the window is diluted and remeasured. The dilution factor then multiplies the diluted concentration back into the original domain. If a sample reads 1.85 AU, dilute one part sample into three parts buffer for a factor of four, read 0.46 AU, and multiply by four to get 1.84 AU. The agreement between the two estimates confirms that the dilute reading was inside the linear window.

## Numbered Walkthrough: A Domain and Range Worksheet From Real Standards

A domain and range worksheet is a table of functions with columns for the rule, the domain, the range, and a sketch. The steps below build one from a six point qPCR standard curve and then verify it against the raw data.

1. Load the standards and fit the line.

```python
import numpy as np
import pandas as pd

standards = pd.DataFrame({
    "copies": [10, 100, 1_000, 10_000, 100_000, 1_000_000],
    "cq":     [35.12, 31.80, 28.55, 25.20, 21.88, 18.60],
})

slope, intercept = np.polyfit(np.log10(standards["copies"]), standards["cq"], 1)
print(f"slope = {slope:.3f}, intercept = {intercept:.3f}")
```

Expected output:

```
slope = -3.306, intercept = 38.429
```

2. State the domain.

```python
print("domain: (0, inf) copies per reaction")
```

Expected output:

```
domain: (0, inf) copies per reaction
```

3. Evaluate the fitted function across the domain and record the modeled range.

```python
grid = np.array([1, 2, 5, 10, 100, 1_000, 1_000_000, 10_000_000], dtype=float)
cq_fit = slope * np.log10(grid) + intercept
for n, c in zip(grid, cq_fit):
    print(f"{n:>12.0f} copies -> Cq {c:6.2f}")
```

Expected output:

```
           1 copies -> Cq  38.43
           2 copies -> Cq  37.43
           5 copies -> Cq  36.12
          10 copies -> Cq  35.12
         100 copies -> Cq  31.82
        1000 copies -> Cq  28.51
     1000000 copies -> Cq  18.59
    10000000 copies -> Cq  15.28
```

4. Compare the modeled range with the observed range.

```python
print(f"observed Cq range: [{standards['cq'].min():.2f}, {standards['cq'].max():.2f}]")
```

Expected output:

```
observed Cq range: [18.60, 35.12]
```

The modeled range runs from 15.28 to 38.43 across the chosen grid. The observed range covers only the six standards. Anything below 18.60 is an extrapolation below the lowest standard, and anything above 35.12 is an extrapolation above the highest.

5. Trigger the domain error on purpose so you recognize it.

```python
import math
math.log10(0)
```

Expected output:

```
Traceback (most recent call last):
  ...
ValueError: math domain error
```

6. Repeat with NumPy to see the vectorized behavior.

```python
np.log10(np.array([0.0, -1.0]))
```

Expected output:

```
RuntimeWarning: divide by zero encountered in log10
RuntimeWarning: invalid value encountered in log10
array([-inf,  nan])
```

7. Fix the input before the function sees it.

```python
raw = np.array([0.0, 12.0, -3.0, 250.0])
clean = raw[raw > 0]
print(clean)
print(np.log10(clean))
```

Expected output:

```
[ 12. 250.]
[1.07918125 2.39794001]
```

That filter is the domain restriction written as code. Apply it before the fit, not after.

8. Save the worksheet.

```python
worksheet = pd.DataFrame({
    "rule": ["Cq = -3.306 log10(N) + 38.429"],
    "domain": ["(0, inf) copies"],
    "range_modeled": [f"[{cq_fit.min():.2f}, {cq_fit.max():.2f}] Cq"],
    "range_observed": [f"[{standards['cq'].min():.2f}, {standards['cq'].max():.2f}] Cq"],
})
print(worksheet.to_string(index=False))
```

Expected output:

```
                              rule      domain range_modeled range_observed
Cq = -3.306 log10(N) + 38.429 (0, inf) copies  [15.28, 38.43] Cq  [18.60, 35.12] Cq
```

## Using a Domain and Range Calculator

A domain and range calculator solves the algebra quickly. It does not know your instrument, your dilution scheme, or the detection limit of your assay.

Workflow:

1. Type the rule with explicit parentheses, for example `log10(x - 3)` rather than `log x - 3`.
2. Select interval notation in the output options.
3. Read the result as a union of intervals.
4. Substitute the endpoints into the original rule to confirm.
5. Compare the algebraic domain with the physical domain of the assay.

| Rule entered | Calculator domain | Calculator range | Physical correction |
| log10(x) | (0, inf) | (-inf, inf) | add the positive concentration ceiling of the assay |
| 1/x | x not equal to 0 | x not equal to 0 | add the non-zero dilution constraint |
| sqrt(x) | [0, inf) | [0, inf) | add the non-negative signal constraint |
| x squared | (-inf, inf) | [0, inf) | input is a measured quantity, so negative values cannot occur |

The last row is the one people miss. The algebra accepts negative inputs, but a negative concentration cannot exist. The mathematical domain is wider than the physical domain, and the report should say which one is being stated.

## Parameters That Matter

Slope. In a standard curve the slope sets how much Cq changes per tenfold change in copy number. A slope near -3.32 corresponds to an efficiency near 100 percent. A slope far from that value widens the uncertainty at the ends of the domain.

Intercept. The intercept sets where the curve crosses one copy. It shifts the whole range up or down and changes the lower limit of useful detection.

EC50. The half maximal concentration. It is a parameter inside the domain, not the domain itself. A reported EC50 above the highest tested dose is a fitted number with no support from the data.

Hill slope. Controls steepness. A Hill slope far from 1 often signals aggregation, competing binding sites, or a fit that has drifted outside the tested range.

Path length. In Beer-Lambert, path length scales the whole function. Doubling the path length doubles the absorbance, so the linear window in concentration shrinks by half.

Blank absorbance. Sets the lower end of the usable absorbance window. A drifting blank moves the lower boundary of the valid range.

Calibration slope for pH. A near ideal electrode gives close to 100 percent of the theoretical Nernst response. A slope far from that value means the electrode needs cleaning, rehydration, or replacement.

## Common Errors and Exact Messages

![Flowchart from raw copy numbers through a zero or negative value to the math domain error and the filter that fixes it.](/article-images/illustrations/domain-and-range-log-domain-error-flow.webp)

| Message | Cause | Fix |
| `ValueError: math domain error` | `math.log10` or `math.sqrt` received an out of domain value | filter inputs to x > 0 before the call |
| `RuntimeWarning: divide by zero encountered in log10` | NumPy received a zero | replace zero with NaN before the log |
| `RuntimeWarning: invalid value encountered in log10` | NumPy received a negative number | clip negatives to NaN, then drop |
| `Warning message: In log(x) : NaNs produced` | base R received zero or a negative input | subset the vector first |
| `#NUM!` | Excel LOG or SQRT received an invalid argument | correct the cell or wrap it in IFERROR |
| `#DIV/0!` | a denominator evaluated to zero | guard the denominator with an IF test |
| EC50 reported outside the tested concentrations | unconstrained fit | constrain the parameter or widen the dose range |

Two failures are silent. A fit that returns flat top and bottom parameters means the tested doses never reached saturation, so the upper end of the range is not supported by the data. A fit that returns an EC50 above the highest dose is reporting a value the experiment never tested. Neither produces an error message.

## How to Check the Result Is Right

Substitute both endpoints into the rule and confirm they reproduce the boundary values. Confirm that the function is monotone over the stated domain if it should be. Check the units of the domain and of the range, because a domain of moles per liter and a range of absorbance units cannot be compared directly. Plot residuals against the input, since a systematic curve in the residuals means the linear domain is narrower than assumed.

For a physical range, compare against an independent bound. In aqueous redox chemistry the thermodynamic stability window of water under standard conditions is 0.000 to 1.229 V, so a predicted redox potential far outside that window is not a useful design target even if the model returns a number [3]. Prediction error also has a scale. Density functional theory estimates of redox potential are reported as differing from experiment by roughly 0.2 to 0.5 V, which is the uncertainty band around the range, not the range itself [3].

For a model-based range, check the applicability domain. QSAR and QSPR models explicitly report an applicability domain, the region of chemical space within which predictions are considered reliable [4]. Outside that region the model is extrapolating and its reported error statistics no longer apply.

## Common Mistakes and Limitations

Treating the codomain as the range. A method that can in principle return any number does not return any number. Report what was observed.

Reporting the calibration domain as the sample domain. If the standards ran from 10 to one million copies, then ten million copies is an extrapolation regardless of what the fit prints.

Extrapolating below the lowest standard. The low end is where the curve flattens and where the log transform is most sensitive to small errors.

Forcing zero onto a log axis. Plot zero dose as a separate control group.

Ignoring saturation. An absorbance above the linear window, a qPCR curve that never plateaus, or a detector at full scale all mean the output has left the valid range.

Confusing the biological domain with the mathematical domain. In structural biology a domain is a [folded region of a protein](/knowledge/bioinformatics/how-to-identify-protein-domains-from-sequence-and-structure-a-step-by-step-guide-using-pfam-and-scop), such as the [conserved catalytic core](/knowledge/bioinformatics/pfam-and-protein-families-how-to-use-domain-annotations-to-enhance-functional-interpretation-of-prot) of acid sphingomyelinases [5]. The same word means two different things, so state which sense you mean.

Forgetting that bounded outputs compress differences. Neural network activation functions with a fixed saturation range compress large inputs into a narrow output band, which is why some architectures add range modulation to recover resolution across the input space [6].

Treating angles and magnitudes as if they spanned the full range. In genome-wide expression analysis, vector algebra represents each expression profile as a vector, and a small set of basis vectors can span the transcriptional range of observed cell behavior [7]. The span is the range of achievable profiles. A new profile far outside that span is not covered by the basis.

Narrowing the spectral domain without saying so. A Raman comparison of chalazion secretions and serum concentrated its differences in the 1100 to 1300 inverse centimeter range, with extrema near 1177, 1184 and 1260 inverse centimeters [8]. The useful domain of that comparison was that band, not the full spectrum.

These rules give the framework for the general case. A specific dataset needs review by the laboratory quality manager and a statistician before results are released.

## Domain and Range Practice With Lab Context

Five items. Answers follow.

1. f(x) = log10(x - 3). State the domain and range.
2. f(x) = sqrt(2x + 6). State the domain and range.
3. A 4PL fit gives bottom = 0.05, top = 1.20, and a positive hill slope. State the domain and range on a log axis.
4. f(x) = 1/(x squared - 4). State the domain and range.
5. Cq = -3.32 log10(N) + 38.5. State the domain and the practical Cq range on a 40 cycle instrument.

Answers.

1. Domain (3, inf). Range (-inf, inf).
2. Domain [-3, inf). Range [0, inf).
3. Domain x > 0. Range (0.05, 1.20).
4. Domain (-inf, -2) U (-2, 2) U (2, inf). Range (-inf, -0.25] U (0, inf).
5. Domain N > 0. Practical Cq range is about 5 to 40 cycles, set at the lower end by the instrument baseline and at the upper end by the cycle limit.

Two extra items for a domain and range worksheet on a plate reader. A dilution series that halves concentration at each step has the domain {c0 / 2 raised to k} for k = 0 to n. That domain is discrete, not an interval. A wavelength scan has a domain given by the instrument start and stop wavelengths, and a range given by the detector's linear response at each wavelength.

## How to Find Domain and Range From a Table of Lab Data

When your data is a table rather than a graph, the domain and range are read directly from the columns. The domain is the set of unique input values listed in the table. The range is the set of unique output values listed next to them.

Steps:

1. Identify the input column and list its unique values in increasing order. That set is the domain.
2. Identify the output column and list its unique values. That set is the range.
3. Note gaps between consecutive inputs. A gap means no measurement was taken there, so those values are not in the domain.
4. Report the observed minimum and maximum of each column.

Worked example:

| Concentration (mg/mL) | Absorbance (AU) |
| 0.0 | 0.002 |
| 0.1 | 0.045 |
| 0.5 | 0.213 |
| 1.0 | 0.441 |
| 2.0 | 0.887 |

Domain: {0.0, 0.1, 0.5, 1.0, 2.0} mg/mL. Range: {0.002, 0.045, 0.213, 0.441, 0.887} AU. Observed span: 0.0 to 2.0 mg/mL and 0.002 to 0.887 AU. There is no data between 0.1 and 0.5 mg/mL, so the domain is not the full interval [0.0, 2.0]; it is the listed set.

Common mistake: treating missing rows as zeros. If a sample at 0.25 mg/mL was not measured, the table has no row for it. Do not insert a row with absorbance 0. A missing row means no observation, and a fabricated zero shrinks the range and biases any fit. Similarly, duplicate wells are listed once; the domain is the set of distinct inputs, not the row count.

## How to Choose the Domain for a Standard Curve

![Decision path for readings above the highest standard, below the lowest standard, or inside the calibration domain.](/article-images/illustrations/domain-and-range-out-of-range-triage.webp)

The domain of a standard curve is the set of standard concentrations you actually measure. Choose it so that every expected unknown falls inside the domain, with at least one standard below the lowest expected value and one above the highest. This avoids extrapolation, where the fitted equation has no data support.

Worked example. You expect unknown samples to contain between 100 and 100,000 copies per reaction. Prepare a 10-fold serial dilution from a 10^6 copies stock: 10^6, 10^5, 10^4, 10^3, 10^2, and 10^1 copies. The usable domain is [10, 10^6] copies. The expected unknowns sit well inside, with one standard below the lowest and one above the highest.

If an unknown reads above the highest standard, dilute it and remeasure. For a 1:10 dilution, multiply the interpolated concentration by 10. For a 1:100 dilution, multiply by 100. If an unknown reads below the lowest standard, report it as below the limit of quantification rather than extrapolating.

Common mistake: choosing evenly spaced concentrations such as 0, 100, 200, 300, 400, 500. On a log scale these bunch at the high end and leave the low end unsampled. Log-spaced standards, each a fixed multiple of the previous one, spread the domain evenly across orders of magnitude and keep the fit stable across the full range. Use 10-fold or 3-fold steps, not linear steps, whenever the assay response is log-linear.

## Domain and Range in Sensor Calibration and Data Logging

A data logger records a raw signal, usually a voltage or a digital count, and converts it to a physical value using a calibration function. The calibration defines the valid domain and range of the entire logging setup. The domain is the span of physical inputs the sensor can measure within specification. The range is the span of raw outputs the logger can represent.

### Clipping and Saturation

Clipping happens when the physical input exceeds the logger's input limits. A 12-bit logger with counts from 0 to 4095 clips any signal that would map below 0 or above 4095. The recorded value sits at the boundary, so the true value is unknown. Saturation is different: the sensor element itself stops responding linearly, even though the logger can still count. Both conditions push data outside the valid domain.

### Worked Example: Rejecting Out-of-Range Readings

A pressure logger is calibrated over 0 to 100 kPa. The calibration line is:

pressure (kPa) = 0.0244 × count

The top count, 4095, corresponds to about 99.9 kPa (0.0244 × 4095), so a count of 4095 could mean about 99.9 kPa or any higher pressure. A count of 0 could mean 0 kPa or a disconnected sensor.

Before analysis, reject every reading at the clipping boundaries:

```python
counts = np.array([0, 512, 4095, 2048, 4095, 1024])
valid = counts[(counts > 0) & (counts < 4095)]
pressures = valid * 0.0244
```

The two clipped readings are removed, leaving four valid pressures. Keeping them would bias the mean downward or upward depending on which boundary they hit. The calibration domain is (0, 100) kPa, and the logger range is (0, 4095) counts. Any reading outside the calibration domain is an extrapolation, not a measurement.

## Related Guides

- [qPCR Data Normalization: Methods and Best Practices](/knowledge/diagnostics/molecular/qpcr-data-normalization-methods-best-practices): Useful for handling Cq data from the qPCR standard curves discussed here
- [Outlier Detection Calculators for Lab Data: A Practical Guide](/blog/data-analysis/outlier-detection-calculator-biology): Helps identify problematic points when defining valid domain and range
- [How to Choose a Descriptive Statistics Calculator for Lab Data](/blog/data-analysis/descriptive-statistics-calculator-for-lab-data): Useful for summarizing observed ranges and distributions in lab data
- [Variance and Standard Deviation Calculators for Lab Data](/blog/data-analysis/sample-variance-calculator): Relevant for understanding spread around fitted ranges and calibration curves
- [EXP Function in Excel: Formula, Examples and Uses](/blog/data-analysis/excel-exp-function-formula-examples): Excel math functions appear in the error table; EXP is a related example
- [Protein Domain Annotation From Sequence Data](/blog/guides/protein-domain-annotation-from-sequence-data): Explains how biological domains are annotated, complementing the protein domain caveat
- [Common Mistakes in Domain Annotation: Why Your Protein Domain Boundaries Are Wrong and How to Fix Them](/knowledge/bioinformatics/common-mistakes-in-domain-annotation-why-your-protein-domain-boundaries-are-wrong-and-how-to-fix-the): Addresses domain boundary errors, extending the article's warning about biological domains
- [How to Plot Protein Domains in Python](/knowledge/bioinformatics/plot-protein-domains-python): Practical for visualizing protein domains if following the Python-based lab workflows

## Frequently Asked Questions

### What is the difference between domain and range?

Domain is the set of inputs a function accepts. Range is the set of outputs it produces.

### How do I write domain and range in interval notation?

Use parentheses for excluded endpoints and square brackets for included ones. Always use a parenthesis next to infinity, as in [0, inf).

### What is the domain of a log function in qPCR?

Copy number must be greater than zero, so the domain is (0, inf). Zero and negative copy numbers are undefined.

### Why does a dose response curve domain exclude zero?

A log concentration axis cannot show zero, and the four parameter logistic model is undefined at x = 0. Zero dose is analyzed as a separate control.

### What is a valid range of pH readings?

A calibrated glass electrode is most reliable in roughly the pH 2 to 12 window. Readings outside the calibration buffers carry more uncertainty.

### What absorbance range is reliable?

Stay inside the instrument's linear window, commonly about 0.1 to 1.0 AU. Dilute and remeasure anything higher.

### Can a domain and range calculator find the physical domain of my assay?

No. It solves the algebra only. You have to add the instrument and sample constraints yourself.

### How do I practice domain and range with lab data?

Fit a standard curve, state its domain, evaluate the endpoints, and compare the observed range with the modeled range.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is the difference between domain and range?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Domain is the set of inputs a function accepts. Range is the set of outputs it produces."
      }
    },
    {
      "@type": "Question",
      "name": "How do I write domain and range in interval notation?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Use parentheses for excluded endpoints and square brackets for included ones. Always use a parenthesis next to infinity, as in [0, inf)."
      }
    },
    {
      "@type": "Question",
      "name": "What is the domain of a log function in qPCR?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Copy number must be greater than zero, so the domain is (0, inf). Zero and negative copy numbers are undefined."
      }
    },
    {
      "@type": "Question",
      "name": "Why does a dose response curve domain exclude zero?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A log concentration axis cannot show zero, and the four parameter logistic model is undefined at x = 0. Zero dose is analyzed as a separate control."
      }
    },
    {
      "@type": "Question",
      "name": "What is a valid range of pH readings?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A calibrated glass electrode is most reliable in roughly the pH 2 to 12 window. Readings outside the calibration buffers carry more uncertainty."
      }
    },
    {
      "@type": "Question",
      "name": "What absorbance range is reliable?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Stay inside the instrument's linear window, commonly about 0.1 to 1.0 AU. Dilute and remeasure anything higher."
      }
    },
    {
      "@type": "Question",
      "name": "Can a domain and range calculator find the physical domain of my assay?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "No. It solves the algebra only. You have to add the instrument and sample constraints yourself."
      }
    },
    {
      "@type": "Question",
      "name": "How do I practice domain and range with lab data?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Fit a standard curve, state its domain, evaluate the endpoints, and compare the observed range with the modeled range."
      }
    }
  ]
}
</script>

## Sources

1. [Sets of range uniqueness for multivariate polynomials and linear functions with rank k.](https://pubmed.ncbi.nlm.nih.gov/36624783/)
2. [A similarity function to evaluate the orthodontic condition in patients with cleft lip and palate.](https://pubmed.ncbi.nlm.nih.gov/15193344/)
3. [RedoxDB and RedEx: closing the DFT-experiment domain gap in redox potential prediction with transfer learning.](https://pubmed.ncbi.nlm.nih.gov/42910327/)
4. [Monte Carlo-based SMILES-driven QSPR modeling of Organic Chromophores across the visible-to- NIR spectral range.](https://pubmed.ncbi.nlm.nih.gov/42594695/)
5. [A conserved fold in a sea of sequence divergence: A theoretical and evolutionary model of the acid sphingomyelinase domain.](https://pubmed.ncbi.nlm.nih.gov/42026436/)
6. [Development of adaptive activation functions with curvature and range modulation for abstract image feature learning in complex CNNs for cross-domain applications.](https://pubmed.ncbi.nlm.nih.gov/42599944/)
7. [Vector algebra in the analysis of genome-wide expression data.](https://pubmed.ncbi.nlm.nih.gov/11897023/)
8. [Visibility-Graph Analysis of Raman Spectra in Chalazion and Lipid Dysregulation: An Exploratory Study.](https://pubmed.ncbi.nlm.nih.gov/42793923/)