Dockerfile

# Use a geospatial-ready R base image
FROM rocker/geospatial:4.4.0

# Install Python 3 and dependencies
RUN apt-get update && apt-get install -y \
    python3-pip \
    python3-dev \
    libosmium2-dev \
    libgdal-dev \
    osmium-tool \
    curl \
    && rm -rf /var/lib/apt/lists/*

# Set working directory
WORKDIR /app

# Set PYTHONPATH to include the src directory
ENV PYTHONPATH="/app/src:${PYTHONPATH}"

# Copy requirement files first for better caching
COPY requirements.txt .

# Install Python packages
RUN pip3 install --no-cache-dir -r requirements.txt

# Install specific R packages needed for the script
RUN R -e "install.packages(c('caret', 'CAST', 'ranger', 'randomForest', 'mapview', 'viridis', 'here', 'doParallel', 'jsonlite', 'pROC', 'yaml'), repos='https://cloud.r-project.org/')"

# Copy the entire project
COPY . .

# Ensure main.py is executable
RUN chmod +x main.py

# Default command: run only training if data is pre-populated, or everything
CMD ["python3", "main.py", "--step", "all"]