SimData <- data.frame( x = runif( 100 ), y = runif( 100 )*100+10 )
SimData$lwr <- SimData$y-10
SimData$upr <- SimData$y+10
xYplot( Cbind( y, lwr, upr )~x, data = SimData )
xYplot( Cbind( y, lwr, upr )~x, data = SimData, scales = list( y = list( log = 10 ) ) )
It seems that error bars are not rescaled (but the points are) with the logarithmic scaling, so the whole thing gets inconsistent.
The quite straightforward xYplot( Cbind( y, log10( lwr ), log10( upr ) )~x, data = SimData, scales = list( y = list( log = 10 ) ) ) solution seems to work, so this is really not a high-priority issue, but perhaps it'd nicer if this'd be automatic.
Here is a minimal reproducible example:
It seems that error bars are not rescaled (but the points are) with the logarithmic scaling, so the whole thing gets inconsistent.
The quite straightforward
xYplot( Cbind( y, log10( lwr ), log10( upr ) )~x, data = SimData, scales = list( y = list( log = 10 ) ) )
solution seems to work, so this is really not a high-priority issue, but perhaps it'd nicer if this'd be automatic.