expert_bigdecimal_precision

🧮 Advanced BigDecimal Precision Control

Ruby’s BigDecimal gives you true arbitrary-precision arithmetic, but by default it uses a modest number of significant digits. You can globally tune precision for all new BigDecimal instances with BigDecimal.limit, and configure rounding modes per calculation via BigDecimal.mode. This is invaluable when you need consistent rounding behavior across complex financial algorithms.

require 'bigdecimal'
require 'bigdecimal/util'

# Set default precision to 30 significant digits
BigDecimal.limit(30)

# Configure rounding mode: HALF_UP for typical financial rounding
BigDecimal.mode(BigDecimal::ROUND_HALF_UP, BigDecimal::ROUND_DIGITS)

# Now every new BigDecimal uses 30 digits and HALF_UP
a = BigDecimal("1.2345678901234567890123456789")
b = BigDecimal("2.3456789012345678901234567890")

sum = a + b
puts sum.to_s("F")  # => high‑precision accurate result