bim-qto

BIM Quantity Takeoff

Overview

Quantity Takeoff (QTO) extracts measurable quantities from BIM models. This skill processes BIM exports to generate grouped quantity reports for cost estimation.

Python Implementation

import pandas as pd
import numpy as np
from typing import Dict, Any, List, Optional, Tuple
from dataclasses import dataclass, field
from enum import Enum


class QTOUnit(Enum):
    """Quantity takeoff measurement units."""
    COUNT = "ea"
    LENGTH = "m"
    AREA = "m2"
    VOLUME = "m3"
    WEIGHT = "kg"
    LINEAR_FOOT = "lf"
    SQUARE_FOOT = "sf"
    CUBIC_YARD = "cy"


@dataclass
class QTOItem:
    """Single QTO line item."""
    category: str
    type_name: str
    description: str
    quantity: float
    unit: str
    level: Optional[str] = None
    material: Optional[str] = None
    element_count: int = 0


@dataclass
class QTOReport:
    """Complete QTO report."""
    project_name: str
    items: List[QTOItem]
    total_elements: int
    categories: int
    generated_date: str


class BIMQuantityTakeoff:
    """Extract quantities from BIM data."""

    # Column mappings for different BIM exports
    COLUMN_MAPPINGS = {
        'type': ['Type Name', 'TypeName', 'type_name', 'Family and Type', 'IfcType'],
        'category': ['Category', 'category', 'IfcClass', 'Element Category'],
        'level': ['Level', 'level', 'Building Storey', 'BuildingStorey', 'Floor'],
        'volume': ['Volume', 'volume', 'Volume (m³)', 'Qty_Volume'],
        'area': ['Area', 'area', 'Surface Area', 'Area (m²)', 'Qty_Area'],
        'length': ['Length', 'length', 'Length (m)', 'Qty_Length'],
        'count': ['Count', 'count', 'Quantity', 'ElementCount'],
        'material': ['Material', 'material', 'Structural Material', 'MaterialName']
    }

    def __init__(self, df: pd.DataFrame):
        """Initialize with BIM data DataFrame."""
        self.df = df
        self.column_map = self._detect_columns()

    def _detect_columns(self) -> Dict[str, str]:
        """Detect which columns exist in data."""
        mapping = {}

        for standard, variants in self.COLUMN_MAPPINGS.items():
            for variant in variants:
                if variant in self.df.columns:
                    mapping[standard] = variant
                    break

        return mapping

    def get_column(self, standard_name: str) -> Optional[str]:
        """Get actual column name from standard name."""
        return self.column_map.get(standard_name)

    def group_by_type(self, sum_column: str = 'volume') -> pd.DataFrame:
        """Group quantities by type name."""

        type_col = self.get_column('type')
        qty_col = self.get_column(sum_column)

        if type_col is None:
            raise ValueError("Type column not found")

        if qty_col is None:
            # Fall back to count
            result = self.df.groupby(type_col).size().reset_index(name='count')
        else:
            result = self.df.groupby(type_col).agg({
                qty_col: 'sum'
            }).reset_index()
            result['count'] = self.df.groupby(type_col).size().values

        result.columns = ['Type', 'Quantity', 'Count'] if len(result.columns) == 3 else ['Type', 'Count']
        return result.sort_values('Count', ascending=False)

    def group_by_category(self, sum_column: str = 'volume') -> pd.DataFrame:
        """Group quantities by category."""

        cat_col = self.get_column('category')
        qty_col = self.get_column(sum_column)

        if cat_col is None:
            raise ValueError("Category column not found")

        agg_dict = {}
        if qty_col:
            agg_dict[qty_col] = 'sum'

        if agg_dict:
            result = self.df.groupby(cat_col).agg(agg_dict).reset_index()
            result['count'] = self.df.groupby(cat_col).size().values
        else:
            result = self.df.groupby(cat_col).size().reset_index(name='count')

        return result.sort_values('count', ascending=False)

    def group_by_level(self, sum_column: str = 'volume') -> pd.DataFrame:
        """Group quantities by building level."""

        level_col = self.get_column('level')
        qty_col = self.get_column(sum_column)

        if level_col is None:
            raise ValueError("Level column not found")

        agg_dict = {}
        if qty_col:
            agg_dict[qty_col] = 'sum'

        if agg_dict:
            result = self.df.groupby(level_col).agg(agg_dict).reset_index()
            result['count'] = self.df.groupby(level_col).size().values
        else:
            result = self.df.groupby(level_col).size().reset_index(name='count')

        return result

    def pivot_by_level_and_type(self) -> pd.DataFrame:
        """Create pivot table: levels as rows, types as columns."""

        level_col = self.get_column('level')
        type_col = self.get_column('type')

        if level_col is None or type_col is None:
            raise ValueError("Level or Type column not found")

        pivot = pd.crosstab(
            self.df[level_col],
            self.df[type_col],
            margins=True
        )

        return pivot

    def filter_by_category(self, categories: List[str]) -> 'BIMQuantityTakeoff':
        """Filter to specific categories."""

        cat_col = self.get_column('category')
        if cat_col is None:
            raise ValueError("Category column not found")

        filtered_df = self.df[self.df[cat_col].isin(categories)]
        return BIMQuantityTakeoff(filtered_df)

    def filter_by_level(self, levels: List[str]) -> 'BIMQuantityTakeoff':
        """Filter to specific levels."""

        level_col = self.get_column('level')
        if level_col is None:
            raise ValueError("Level column not found")

        filtered_df = self.df[self.df[level_col].isin(levels)]
        return BIMQuantityTakeoff(filtered_df)

    def get_walls(self) -> pd.DataFrame:
        """Get wall quantities."""
        cat_col = self.get_column('category')
        if cat_col:
            walls = self.df[self.df[cat_col].str.contains('Wall', case=False, na=False)]
            return BIMQuantityTakeoff(walls).group_by_type()
        return pd.DataFrame()

    def get_floors(self) -> pd.DataFrame:
        """Get floor/slab quantities."""
        cat_col = self.get_column('category')
        if cat_col:
            floors = self.df[self.df[cat_col].str.contains('Floor|Slab', case=False, na=False)]
            return BIMQuantityTakeoff(floors).group_by_type()
        return pd.DataFrame()

    def get_doors(self) -> pd.DataFrame:
        """Get door quantities."""
        cat_col = self.get_column('category')
        if cat_col:
            doors = self.df[self.df[cat_col].str.contains('Door', case=False, na=False)]
            return BIMQuantityTakeoff(doors).group_by_type()
        return pd.DataFrame()

    def get_windows(self) -> pd.DataFrame:
        """Get window quantities."""
        cat_col = self.get_column('category')
        if cat_col:
            windows = self.df[self.df[cat_col].str.contains('Window', case=False, na=False)]
            return BIMQuantityTakeoff(windows).group_by_type()
        return pd.DataFrame()

    def generate_report(self, project_name: str = "Project") -> QTOReport:
        """Generate complete QTO report."""

        from datetime import datetime

        items = []
        type_col = self.get_column('type')
        cat_col = self.get_column('category')
        level_col = self.get_column('level')
        vol_col = self.get_column('volume')
        area_col = self.get_column('area')
        mat_col = self.get_column('material')

        # Group by type
        grouped = self.df.groupby(type_col if type_col else self.df.columns[0])

        for type_name, group in grouped:
            # Determine primary quantity
            qty = 0
            unit = QTOUnit.COUNT.value

            if vol_col and vol_col in group.columns:
                qty = group[vol_col].sum()
                unit = QTOUnit.VOLUME.value
            elif area_col and area_col in group.columns:
                qty = group[area_col].sum()
                unit = QTOUnit.AREA.value
            else:
                qty = len(group)
                unit = QTOUnit.COUNT.value

            # Get category and material
            category = group[cat_col].iloc[0] if cat_col and cat_col in group.columns else ""
            material = group[mat_col].iloc[0] if mat_col and mat_col in group.columns else ""
            level = group[level_col].iloc[0] if level_col and level_col in group.columns else ""

            items.append(QTOItem(
                category=str(category),
                type_name=str(type_name),
                description=str(type_name),
                quantity=round(qty, 2),
                unit=unit,
                level=str(level) if level else None,
                material=str(material) if material else None,
                element_count=len(group)
            ))

        return QTOReport(
            project_name=project_name,
            items=items,
            total_elements=len(self.df),
            categories=self.df[cat_col].nunique() if cat_col else 0,
            generated_date=datetime.now().isoformat()
        )

    def to_excel(self, output_path: str, project_name: str = "Project"):
        """Export QTO to Excel with multiple sheets."""

        with pd.ExcelWriter(output_path, engine='openpyxl') as writer:
            # Summary by category
            self.group_by_category().to_excel(
                writer, sheet_name='By Category', index=False)

            # Summary by type
            self.group_by_type().to_excel(
                writer, sheet_name='By Type', index=False)

            # Level breakdown
            try:
                self.pivot_by_level_and_type().to_excel(
                    writer, sheet_name='Level-Type Matrix')
            except:
                pass

            # Walls
            walls = self.get_walls()
            if not walls.empty:
                walls.to_excel(writer, sheet_name='Walls', index=False)

            # Doors and Windows
            doors = self.get_doors()
            if not doors.empty:
                doors.to_excel(writer, sheet_name='Doors', index=False)

            windows = self.get_windows()
            if not windows.empty:
                windows.to_excel(writer, sheet_name='Windows', index=False)

        return output_path