chart

Chart

Use nong for statistical analysis and implemented figure generation. Nong.Toolkit.Net routes to the CLI; do not recreate chart rendering logic in scripts or temporary projects.

Nong CLI Preflight

Read ../../.claude/references/nong-cli-preflight.md before the first Nong CLI command in a session. Confirm the nong CLI is installed and the chart command group.

Modular: nong chart routes to the standalone Angri450.Nong.Tool.Chart dotnet tool. First use auto-installs via dotnet tool install --global. Command surface unchanged.

Implemented Commands

nong chart analyze <groups.json> [--alpha 0.05] [--json]
nong chart anova <groups.json> [--json]
nong chart duncan <groups.json> [--alpha 0.05] [--json]
nong chart bar <groups.json> -o <out.png> [--title <text>] [--ylabel <text>] [--error sem|none] [--no-significance] [--json]
nong chart line <spec.json> -o <out.png> [--json]
nong chart scatter <spec.json> -o <out.png> [--json]
nong chart pie <spec.json> -o <out.png> [--json]
nong chart boxplot <groups.json> -o <out.png> [--title <text>] [--ylabel <text>] [--json]
nong chart histogram <values.json> -o <out.png> [--title <text>] [--xlabel <text>] [--ylabel <text>] [--bin-count 20] [--json]
nong chart heatmap <spec.json> -o <out.png> [--title <text>] [--colormap <name>] [--json]
nong chart radar <spec.json> -o <out.png> [--title <text>] [--json]

Dispatch

1. For one-shot agricultural treatment analysis, run nong chart analyze <groups.json> --json.
2. For ANOVA only, run nong chart anova <groups.json> --json.
3. For Duncan multiple comparison only, run nong chart duncan <groups.json> --json.
4. For a publication-style bar chart with optional significance letters, run nong chart bar <groups.json> -o <out.png> --json.
5. For trends, use nong chart line <spec.json> -o <out.png> --json.
6. For x/y relationships, use nong chart scatter <spec.json> -o <out.png> --json.
7. For composition shares, use nong chart pie <spec.json> -o <out.png> --json.
8. For treatment group distribution comparison, use nong chart boxplot <groups.json> -o <out.png> --json.
9. For data distribution visualization, use nong chart histogram <values.json> -o <out.png> --json.
10. If data starts in Excel, use excel to-groups --raw first for analysis/bar charts.
11. For treatment group distribution, use nong chart heatmap <spec.json> -o <out.png> --json (heatmap input: {"data":[[...]],"rows":N,"cols":M}).
12. For multi-index comparison, use nong chart radar <spec.json> -o <out.png> --json (radar input: {"categories":["A","B"],"series":[{"name":"X","values":[1,2]}]}).
13. Do not promise combined panels or figure editing unless a future nong commands --json exposes those as implemented CLI commands.

Input Contracts

Grouped JSON for analyze, anova, duncan, and bar:

{
  "Control": [1.2, 1.3, 1.1],
  "Treatment": [2.0, 2.2, 2.1]
}

Line chart spec:

{
  "title": "Growth",
  "xLabel": "Days",
  "yLabel": "Height",
  "series": [
    { "name": "A", "x": [0, 7, 14], "y": [1.0, 2.0, 3.0] }
  ]
}

Scatter plot spec:

{
  "title": "Correlation",
  "xLabel": "pH",
  "yLabel": "Yield",
  "points": [
    { "x": 6.1, "y": 12.3, "group": "A" }
  ],
  "trendline": true
}

Pie chart spec:

{
  "title": "Composition",
  "values": [
    { "label": "A", "value": 30 },
    { "label": "B", "value": 70 }
  ]
}

Treat E006 validation_failed as a data or spec problem to fix before analysis or rendering.

Optional Visual QA

After generating a PNG, you may suggest:

nong ocr analyze-image fig.png -o fig.analysis --json

This is structural image QA for dimensions, blankness, whitespace, and content regions. It is not OCR, text recognition, or semantic understanding of the figure.