Data and Plot Settings
Formula Used
A dot plot places one marker for every observation. Repeated values stack vertically to show frequency. Grouped modes separate classes, models, or prediction outcomes.
Mean = Σx / n Sample variance = Σ(x − x̄)² / (n − 1) Sample standard deviation = √variance IQR = Q3 − Q1 Possible outlier: x < Q1 − 1.5×IQR or x > Q3 + 1.5×IQR Accuracy = correct predictions / evaluated predictions × 100
How to Use
- Paste numeric values or structured machine learning rows.
- Select a preset and suitable dot plot style.
- Choose filters, transformations, labels, and display controls.
- Generate the plot and review statistical summaries.
- Export the chart, table, data, or configuration.
Example Data Format
| value | category | actual | predicted | confidence | model |
|---|---|---|---|---|---|
| 2.1 | Class A | A | A | 0.88 | Model One |
| 3.4 | Class B | B | A | 0.55 | Model One |
| 4.1 | Class C | C | C | 0.97 | Model Two |
Frequently Asked Questions
What data can this calculator accept?
It accepts numeric lists and structured CSV-style rows. Optional fields support categories, labels, predictions, confidence, and models. Invalid numeric rows are skipped safely.
How are repeated values displayed?
Stacked mode places repeated values above each other. Their vertical position represents frequency. The frequency table reports exact totals.
Can it compare machine learning classes?
Yes, category fields create separate datasets. Presets can compare models or prediction outcomes. Class summaries report counts and means.
How are outliers detected?
The calculator uses the 1.5 IQR rule. Values outside both fences are highlighted. These remain possible outliers, not automatic errors.
Can I analyze prediction accuracy?
Provide actual and predicted columns together. The calculator marks correct and incorrect rows. It then calculates overall accuracy.
What does the confidence threshold do?
It separates scores meeting your chosen threshold. This supports confidence distribution analysis. It does not change predicted labels.
What transformations are available?
You can normalize, standardize, or apply logarithms. Logarithms require positive observations. Results preserve original values for reference.
Can I remove duplicates?
Yes, duplicate removal considers value and ML fields. It is optional. Keep duplicates when frequency is meaningful.
Are uploaded files stored?
This page processes files during the request. It does not intentionally save uploaded datasets. Server policies may still apply.