Pair Plot Calculator

Build interactive pair plot matrices, compare feature distributions, reveal correlations, detect outliers, explore grouped patterns, and export clear machine learning visual insights instantly online.

1. Data input

No dataset parsed.

2. Feature and grouping selection

Choose two to eight features for responsive performance.

3. Data preprocessing

Zero keeps all rows.

4. Statistical and plot options

Zero selects automatically.

5. Appearance and axes

Formula used

Pearson correlation: covariance divided by both standard deviations. Values near one show strong positive movement.

Spearman correlation: Pearson correlation applied to ranked values. It detects monotonic relationships with fewer distribution assumptions.

Kendall correlation: concordant pairs minus discordant pairs, divided appropriately. It measures ordinal agreement between two variables.

IQR outliers: values beyond Q1 − k×IQR or Q3 + k×IQR. Z-score outliers exceed the selected standard-deviation threshold.

How to use

  1. Paste data, upload CSV, or enter a manual table.
  2. Parse the dataset and inspect detected column types.
  3. Select two through eight numeric features.
  4. Choose grouping, preprocessing, statistics, and visual settings.
  5. Generate the matrix and inspect relationships.
  6. Export charts, statistics, processed data, or settings.

Example data

sepal_lengthsepal_widthpetal_lengthpetal_widthspecies
5.13.51.40.2Setosa
6.43.24.51.5Versicolor
6.33.36.02.5Virginica

Frequently asked questions

What does a pair plot show?

It shows every selected feature against every other feature. Diagonal cells display individual feature distributions.

How many features should I select?

Four to six features usually remain easy to inspect. Eight features are supported for broader exploration.

Does correlation prove causation?

No. Correlation only measures statistical association between variables. Domain knowledge remains essential.

When should I use Spearman correlation?

Use it for monotonic relationships or ranked data. It handles strong nonlinearity better than Pearson.

Why can outliers change correlations?

Extreme values can strongly pull fitted relationships. Compare results before excluding detected outliers.

What does standardisation do?

It centres each feature around zero. It scales values using their standard deviation.

What does grouped plotting provide?

Groups receive separate markers and colours. This reveals class-specific clusters and relationships.

Why are some rows removed?

Invalid transformations or missing values may remove rows. Warnings explain each preprocessing decision.

Can I export processed data?

Yes. Download the transformed dataset and correlation matrix as separate CSV files.

Related Calculators

Scatter PlotBubble ChartScatterplot MatrixCorrelation HeatmapHexbin Plot2D Density PlotJoint PlotCategorical Scatter PlotParallel Coordinates PlotAndrews Curves

Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.