Skip to contents

Convert Integer Columns to Numeric (Optimized)

Source: R/analyze_employment_transitions.R
dot-convert_types_optimized.Rd

Efficiently converts all integer columns in a data.table to numeric type to prevent type conflicts during consolidation operations. Uses vectorized operations for maximum performance.

Usage

.convert_types_optimized(data, modify_in_place = FALSE)

Arguments

data

data.table object to process

modify_in_place

Logical; if TRUE, modifies data in place using :=. If FALSE (default), creates a copy first to preserve original data.

Value

data.table with integer columns converted to numeric

Details

This optimization addresses memory and performance issues caused by:

  • Multiple redundant type checking loops across the codebase

  • Type conflicts during data.table aggregation operations

  • Unnecessary data copies for type conversions

The function uses vectorized sapply() to identify integer columns in a single pass, then batch-converts them to numeric using data.table's efficient := operator. This approach is significantly faster than iterating through columns individually, especially for wide datasets.

Memory efficiency:

  • With modify_in_place=TRUE: Zero additional memory allocation

  • With modify_in_place=FALSE: Single copy() operation

  • Replaces 3+ redundant copies in .apply_temporal_consolidation()

  • Eliminates repeated type checking loops (lines 778-825, 4047-4058)