Why did some numeric values turn into null even though Excel shows numbers?

FAQs
1

Sparkflows never converts invalid numeric values into strings once a column is inferred as numeric.

With the updated engine:

  • Valid numeric formats are parsed correctly

  • Percentages, accounting formats, and grouped numbers are normalized

  • Invalid numeric representations become null

  • The column type is preserved

Examples that become null:

  • Invalid comma placement

  • Mixed symbols in numeric columns

  • Formula errors

  • Text embedded in numeric data

Why this is intentional

Silently converting bad numeric values to strings causes:

  • Broken aggregations

  • Incorrect joins

  • Downstream data corruption

Sparkflows prefers correct schema with nulls over incorrect data.

Recommendation

If nulls are unexpected:

  • Inspect the original Excel formatting

  • Or enforce schema explicitly for full control