On 2021-Sep-12, Dean Rasheed wrote: > So the fix is just to remove the upper bound on this local_rscale, as > we do for the full-precision calculation. This doesn't impact > performance, because it's only computing the logarithm to 8 > significant digits at this stage, and when x is very close to 1 like > this, ln_var() has very little work to do -- there is no argument > reduction to do, and the Taylor series terminates on the second term, > since 1-x is so small.
I came here just to opine that there should be a comment about there not being a clamp to the maximum scale. For example, log_var says "Set the scales .. so that they each have more digits ..." which seems clear enough; I think the new comment is a bit on the short side. > Coming up with a test case that doesn't have thousands of digits is a > bit fiddly, so I chose one where most of the significant digits of the > result are a long way after the decimal point and shifted them up, > which makes the loss of precision in HEAD more obvious. The expected > result can be verified using bc with a scale of 2000. I couldn't get bc (version 1.07.1) to output the result; it says Runtime warning (func=(main), adr=47): non-zero scale in exponent Runtime error (func=(main), adr=47): exponent too large in raise -- Álvaro Herrera PostgreSQL Developer — https://www.EnterpriseDB.com/