nhs-r-community/FunnelPlotR

Limits calculation not matching SHMI

chrismainey opened this issue · 2 comments

Tau and Phi seem to be a bit out on the most recent version of SHMI, making limits different. Reported by @richardi7 originally.

Fixed several issues in 7996ecf related to the NHSD SHMI methods that use truncation instead of winsorisation:

  • bug in truncation step that only truncated one side of distribution.
  • NHSD appear to use on two decimal places on on expected. needs further checking, as publication is censored.
  • Didn't affect all other methods, but needed to exclude the truncated S values from the Tau2 calculation.

Needs further testing.

Resolved, minor fix going to CRAN today.