Skip to main content

Example: Data Analysis with Polars

Demo: Collect Revit data → Analyze with Polars → Visualize results Script: data_analysis_script.py Time: 2-3 minutes

What It Does​

Complete workflow combining all three modules:

  1. Code Execution - Auto-install Polars via PEP 723
  2. Logging - Structured output with syntax highlighting
  3. Visualization - Display outliers in 3D view

Key Pattern: PEP 723 Dependencies​

# /// script
# dependencies = [
#     "polars==1.38.1",
# ]
# ///

When you click Execute:

  • Code Execution checks if polars is installed
  • If missing, Pixi installs it automatically (~5 seconds). On Plant 3D the same metadata uses the uv sidecar instead.
  • Script runs with full access to Polars DataFrame API

Key Pattern: Revit Data Collection​

from Autodesk.Revit.DB import FilteredElementCollector, Wall

doc = __revit__.ActiveUIDocument.Document
walls = FilteredElementCollector(doc).OfClass(Wall).ToElements()

data = []
for wall in walls:
    data.append({
        "Id": wall.Id.IntegerValue,
        "Name": wall.Name,
        "Area": wall.get_Parameter(BuiltInParameter.HOST_AREA_COMPUTED).AsDouble(),
    })

Output in Trace Panel:

Collecting wall data...
Collected 127 walls

Key Pattern: DataFrame Analysis​

import polars as pl

df = pl.DataFrame(data)

# Statistical analysis
stats = df.select(pl.col("Area").describe())
print(stats)

# Find outliers (area > 2 std deviations)
mean_area = df["Area"].mean()
std_area = df["Area"].std()
outliers = df.filter(pl.col("Area") > mean_area + 2 * std_area)

Output in Trace Panel:

Wall Area Statistics:
┌──────────┬────────────┐
│ statistic│ value      │
├──────────┼────────────┤
│ count    │ 127.0      │
│ mean     │ 245.8      │
│ std      │ 78.3       │
│ min      │ 12.5       │
│ max      │ 892.1      │
└──────────┴────────────┘

Found 3 outlier walls

Key Pattern: Geometry Visualization​

# Get walls from outlier IDs
outlier_ids = [ElementId(id) for id in outliers["Id"]]
outlier_walls = [doc.GetElement(id) for id in outlier_ids]

# Print geometry → auto-visualized in 3D
for wall in outlier_walls:
    curve = wall.Location.Curve
    print(f"Wall {wall.Id}: Area={wall_area:.1f} sq ft")
    print(curve)  # Appears in 3D view with color

Result:

  • Outlier walls highlighted in 3D view
  • Trace Panel shows IDs + areas
  • Clear visual feedback for analysis

Try It Yourself​

  1. Open RevitDevTool panel
  2. Navigate to Code Execution tab
  3. Load folder: samples/PythonDemo/commands/
  4. Click data_analysis_script.py
  5. Watch dependencies install (first run only)
  6. See results in Trace Panel + 3D view

Full source: data_analysis_script.py