Skip to content

Running Analysis

This guide explains how to run population analysis using the pypopRF plugin and interpret the results.

Before Starting

Quick Checklist

  • ✓ All required files loaded
  • ✓ Optional files prepared (if using)
  • ✓ Settings configured
  • ✓ Sufficient disk space (3-4x input data size)
  • ✓ Adequate memory available
  • ✓ Column names verified

Data Verification

  • Census data columns match settings
  • Mastergrid IDs match census data
  • Covariates cover study area
  • Optional files properly configured:
  • Water mask (if using)
  • Constraints (if using)
  • Age-sex data (if using)

Running the Analysis

1. Starting the Process

The Start button will be enabled when:

  • All required files are loaded
  • Settings are properly configured
  • Project is initialized

Click the green "Start" button to begin analysis.

2. Processing Stages

The analysis runs through several stages:

Feature Extraction (20%)

  • Processing covariates
  • Calculating zonal statistics
  • Extracting predictive features

Model Training (40%)

  • Training Random Forest model
  • Calculating feature importance
  • Validating model performance

Population Prediction (60%)

  • Generating probability surface
  • Applying water mask (if enabled)
  • Processing spatial constraints (if enabled)

Distribution Mapping (80%)

  • Creating normalized census
  • Generating population distribution
  • Processing age-sex structure (if enabled)

Finalization (100%)

  • Saving output files
  • Loading results in QGIS
  • Cleaning temporary files

Monitoring Progress

Console Output

The console shows:

  • Current processing stage
  • Progress information
  • Warnings or errors
  • Processing statistics

Progress Bar

  • Shows overall completion
  • Displays current stage
  • Updates in real-time

Process Control

  • Use Stop button to cancel (red)
  • Processing will complete current stage
  • Results up to stop point are saved

Output Files

Main Outputs

  1. normalized_census.tif

    • Census values adjusted to raster format
    • Shows population distribution zones

  2. population_unconstrained.tif

    • Basic population distribution
    • Without spatial constraints

  3. population_constrained.tif

    • Population distribution with constraints
    • Only if constraints enabled

Additional Outputs

Age-sex maps (if enabled)

  • Separate maps for each age-sex group
  • Located in 'agesex' folder

Model files

  • model.pkl.gz: Trained model
  • scaler.pkl.gz: Data scaler
  • features.csv: Feature importance