DrawIO Non-Data Diagram Creation
This skill follows
. It is responsible only for
non-data diagrams in the paper, such as technical roadmaps, solution flowcharts, model structure diagrams, data processing flowcharts, variable relationship diagrams, indicator system diagrams, and so on.
Mathematical Modeling Standards Reference
If domain judgment is needed, read the “Charts and Visualization” and “Non-Data Diagram Tool Selection” sections in
../_references/math_modeling_norms.md
. This file serves only as a standards knowledge base; do not generate extra diagrams merely to increase the count.
Stage Boundaries
- This stage is responsible for: DrawIO source files, non-data diagram PDFs, and diagram generation records.
- This stage is not responsible for: line charts, bar charts, scatter plots, heatmaps, box plots, radar plots, and other data-driven charts. These are generated by .
- This stage does not rerun models, modify , or rewrite the numerical conclusions in
reports/RESULTS_REPORT.md
.
Required Outputs
Create or update the following files in the current working directory:
text
figures/
fig_roadmap.drawio
fig_roadmap.pdf
fig_flow_q1.drawio
fig_flow_q1.pdf
...
reports/DRAWIO_REPORT.md
If a particular type of diagram is not needed, explain the reason in
. Do not treat the technical roadmap as a fixed-count requirement; generate it only if it can communicate genuine methodological relationships.
The purpose of reading these files is not to extract data for charting, but to understand the paper’s methodology, chapter structure, subproblem relationships, and existing figures, thereby avoiding duplication.
Workflow
Step 1: Inventory Existing Figures and Requirements
First read the following files if they exist:
reports/ANALYSIS_MODELING_REPORT.md
,
reports/RESULTS_REPORT.md
, and the
directory listing.
Then extract the non-data diagram requirements from the preceding documents and output a checklist:
text
DRAWIO PLAN CHECKLIST:
[ ] fig_roadmap Technical roadmap, placed in the Problem Restatement/Introduction
[ ] fig_flow_q1 Problem 1 solution flowchart
[ ] fig_flow_q2 Problem 2 solution flowchart
[ ] fig_flow_q3 Problem 3 solution flowchart
[ ] fig_pipeline Data processing flowchart
[ ] fig_model Model structure/variable relationship diagram
The checklist is not a fixed template. Adjust it according to the actual problem by removing or adding items. Do not generate meaningless diagrams just to increase the count.
Step 2: Determine the Diagram Type
Common diagram choices:
| Diagram Type | Suggested Filename | Use Case |
|---|
| Technical roadmap | | Present the overall solution route, chapter logic, and method sequence |
| Subproblem solution flowchart | , | Present the inputs, decisions, algorithms, and outputs of an individual subproblem |
| Data processing flowchart | | Present data cleaning, feature construction, and model inputs |
| Model structure diagram | | Present module relationships, variable relationships, and model hierarchy |
| Indicator system diagram | | Present the objective, criteria, and indicator layers |
| Decision tree/rule diagram | | Present classification rules, equipment selection, or strategy branches |
Do not use DrawIO for the following:
- Result comparison bar charts
- Prediction error curves
- Sensitivity curves
- Correlation heatmaps
- Distribution plots and box plots
Step 3: Generate DrawIO Source Files
Create one
file per diagram and place it in
.
DrawIO content requirements:
- Use the same text language as the paper.
- Keep node text short, using two lines if necessary; do not stack long sentences.
- Use consistent styles for nodes of the same type.
- Keep arrow directions clear and avoid crossings.
- Do not place lengthy explanations in the diagram; leave them in the paper’s main text.
- Do not use decorative shadows or excessive gradients.
When generating a large XML file, use a reliable file-editing approach that writes it in sections and then verify that the closing tags are complete to avoid truncation.
Step 4: Export PDFs
Prefer an available DrawIO command to export PDFs:
bash
DRAWIO_BIN="$(command -v drawio 2>/dev/null || command -v draw.io 2>/dev/null || command -v draw.io.exe 2>/dev/null || true)"
if [ -n "$DRAWIO_BIN" ]; then
"$DRAWIO_BIN" --export --format pdf --crop --output figures/fig_roadmap.pdf figures/fig_roadmap.drawio
else
echo "DrawIO command not found; keep .drawio source and record export failure."
fi
If PDF export is unavailable, retain the
file and record the failure reason and the recommended export command in
.
Step 5: Self-Inspection and Repair
Every diagram must be checked for the following:
- The file is not empty.
- If the export succeeds, the file is not empty.
- Nodes do not visibly overlap.
- Arrows do not pass through core nodes.
- Font sizes, colors, and border styles are consistent.
- The filename matches the diagram’s meaning.
- The diagram does not duplicate a data chart from .
If any issue is found, fix the
file and export it again. Do not merely explain the issue in the report.
Step 6: Write the Generation Record
Create
containing at least the following:
markdown
# DrawIO Diagram Generation Report
## Diagram Inventory
| --- | --- | --- | --- | --- |
## Ungenerated Diagrams and Reasons
## Export and Self-Inspection Log
## Embedding Recommendations for the Paper Stage
In CUMCM 2026 mode, the diagram inventory must also record whether it contains an AI-generated structure and the manual inspection status. Append this stage’s AI involvement to
compliance/AI_USAGE_LOG.md
. Do not represent an incomplete manual inspection as “verified.”
Embedding recommendations should only identify the chapter appropriate for each diagram and suggest a caption. Do not generate
. The final figure insertion code, whether Typst’s
#figure(image(...), caption: [...])
or LaTeX’s
\begin{figure}...\end{figure}
, is determined by
based on the paper structure and selected engine.
Quality Requirements
- Diagrams should support the paper’s argumentation rather than serve as decoration.
- Every diagram must correspond to a real method in
reports/ANALYSIS_MODELING_REPORT.md
.
- Data-driven charts must not be regenerated during this stage.
- Every non-data diagram cited during the paper stage must have both a source file and a PDF, or the export failure must be explained in .