Reports & Visualization

CodeGreen provides multiple output formats and an interactive energy timeline visualization.

Output Formats

Console Output (default)

codegreen measure python script.py -g fine

Prints total energy, wall time, and per-function breakdown to the terminal.

JSON Output

codegreen measure python script.py --json

Machine-readable JSON with analysis metadata, measurement results, and checkpoint data.

CSV Output

codegreen measure python script.py --csv

Markdown Output

codegreen measure python script.py --markdown

File Output

codegreen measure python script.py -o results.json

Energy Timeline Visualization

N-body simulation (50,000 iterations): advance identified as the energy hotspot at 17.21J. Timeline shows linear energy ramp during computation.

Interactive HTML (Plotly)

codegreen measure python script.py -g fine --export-plot energy.html

Generates a self-contained HTML file with:

• Function Energy Bar Chart: Horizontal bars sorted by energy. Hotspots (>90^th percentile) in red.
• Zoomable Timeline: Energy vs time scatter plot. Green = enter, red = exit. Supports scroll zoom, drag pan, box select, hover tooltips.
• Summary Statistics: Total energy (J), wall time, average/peak power (W), checkpoint count.

Works offline -- Plotly.js is embedded in the file.

For large checkpoint counts (100K+), hotspot-aware downsampling preserves energy-intensive functions.

Static Image (matplotlib)

codegreen measure python script.py -g fine --export-plot energy.png

Dual-panel image: function energy bars (top) + timeline scatter (bottom). Requires pip install matplotlib.

Dependencies

Format	Dependency	Install
.html	plotly	pip install plotly
.png / .pdf	matplotlib	pip install matplotlib

Or: pip install -e ".[viz]"

Reading the Results

• Energy (J): Cumulative energy from RAPL counter
• Delta (J): Energy between consecutive checkpoints
• Power (W): Instantaneous power at each checkpoint
• Hotspots: Functions exceeding 90^th percentile energy