Number

The Number column type lets you store numeric values in your Phacet: integers or decimals. Values can be entered manually, extracted from documents using AI, or computed with Python code.

Unlike Text columns, Number columns enforce a strict numeric format: only valid numbers (or empty values) are accepted. This ensures data consistency and makes your columns ready for calculations, comparisons, and aggregations.

Example use cases

• Extract financial figures: pull revenue, costs, or margins from contracts, invoices, or financial reports.
• Score or rank records: assign numeric scores to leads, applications, or risk assessments.
• Capture measurements: store quantities, durations, distances, or any measurable attribute from source documents.
• Compute derived values: calculate totals, averages, ratios, or deltas using Python across multiple columns.

Setting Up a Number Column

Step 1: Create the column

• Add a new column to your Phacet.
• Choose Number as the column type.
• Give it a clear, descriptive name (e.g. Revenue, Unit Price, Risk Score).

Step 2: Select your tool

Decide how the numeric values will be populated:

• Manual Input: enter numbers directly into each cell.
• AI (LLM): AI will analyze input data and extract numeric values automatically.
• Python Tool: compute values using custom logic, formulas, or rules.

Manual vs Automated input

Referencing other data

When using the AI Tool or Python Tool, you can reference data from other columns using the @ symbol followed by the column name (e.g. @Invoice Document, @Unit Price).

This allows your tool to use the value from that column as input.

AI Tool setup

• Write clear instructions describing the numeric value to extract.
• Reference the source column containing the data to analyze (e.g. a File column or a Text column).
• AI will return a numeric value based on its interpretation of the content.

Learn more about the AI tool in the dedicated article

Python Tool setup

• Your Python main() function must return a numeric value (int or float) or None if no value applies.
• Returning a string, list, or any other type will cause a validation error.

Learn more about the Python tool in the dedicated article

Example prompts and scripts

AI Tool

Extract revenue from a financial report (single value)

Extract the total annual revenue in euros from @Financial Report. Return only the numeric value, without currency symbols or units.

Extract a quantity from an invoice

Read @Invoice document and extract total amount including taxes.

Python Tool

Compute a margin from two columns

def main():
    revenue = @Revenue
    cost = @Cost
    if revenue is None or cost is None:
        return None
    return revenue - cost

Apply a percentage

def main():
    base_price = @Unit Price
    if base_price is None:
        return None
    return round(base_price * 1.20, 2)

Convert text to number

def main():
    raw = @Raw Value
    if raw is None:
        return None
    cleaned = raw.replace(" ", "").replace(",", ".")
    return float(cleaned)

Number vs Text for numeric data

You may be tempted to store numeric values in a Text column. Here's why Number columns are a better fit:

Best practices

• Use descriptive column names: Include the unit or context when relevant (e.g. Revenue (EUR), Weight (kg)).
• Handle missing values gracefully: Always account for None in your Python functions to avoid runtime errors.
• Keep precision in mind: Use round() in Python when dealing with decimals to avoid floating-point artifacts.
• Validate at the source: If extracting numbers with AI, include explicit instructions about format expectations (no currency symbols, no thousand separators) to get clean results.
• Combine with other column types: Use Number columns alongside Text or Select columns to build complete data processing pipelines: extract with AI, compute with Python, classify with Select.