lenthomas
commented
5 years ago I would think the bootstrap mean is customary (it's supposed to be less biased than the mean of the estimate before bootstrapping), but it is susceptible to outliers, so it might be wise to produce both, with some documentation about them.
Distance for windows does not produce either - it just gives the original point estimate, with the bootstrap just used to produce a bootstrap SE, with log-normal confidence intervals, and a bootstrap percentile-based CI. So that might be an option, making the bootstrap point estimate an option (with the user specifying which they want).
Len
On 20-Aug-19 8:44 AM, erex wrote:
Don't expect this to get sorted before tomorrow, but there is scant documentation for this function. Breadcrumbs in existing documentation refer to sections that don't exist as here: https://github.com/DistanceDevelopment/Distance/blob/17a02337de5ea2979adef2a4df5fd7c36a652b99/R/bootdht.R#L12
What got my attention was that the |summary()| function spits out the median of the replicate estimates. I don't know if that is customary (rather than the mean), but the output is not labelled as median, so I looked to documentation which is silent on the subject.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/DistanceDevelopment/Distance/issues/33?email_source=notifications&email_token=AAA7T36RXBJBDUJWEJXVGDTQFOOGDA5CNFSM4INRAAW2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HGFMKNQ, or mute the thread https://github.com/notifications/unsubscribe-auth/AAA7T36NLJEFFSHV72QYMHTQFOOGDANCNFSM4INRAAWQ.Web Bug from https://github.com/notifications/beacon/AAA7T3ZFKKSTBMYONL2VHV3QFOOGDA5CNFSM4INRAAW2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HGFMKNQ.gif
[ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/DistanceDevelopment/Distance/issues/33?email_source=notifications\u0026email_token=AAA7T36RXBJBDUJWEJXVGDTQFOOGDA5CNFSM4INRAAW2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HGFMKNQ", "url": "https://github.com/DistanceDevelopment/Distance/issues/33?email_source=notifications\u0026email_token=AAA7T36RXBJBDUJWEJXVGDTQFOOGDA5CNFSM4INRAAW2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HGFMKNQ", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]
-- Len Thomas len.thomas@st-andrews.ac.uk lenthomas.org @len_thom Centre for Research into Ecological and Environmental Modelling The Observatory, University of St Andrews, Scotland KY16 9LZ Office: UK+1334-461801 Admin: UK+1334-461842
The University of St Andrews is a charity registered in Scotland, No SC013532.
dill
erex
Don't expect this to get sorted before tomorrow, but there is scant documentation for this function. Breadcrumbs in existing documentation refer to sections that don't exist as here: https://github.com/DistanceDevelopment/Distance/blob/17a02337de5ea2979adef2a4df5fd7c36a652b99/R/bootdht.R#L12
What got my attention was that the
summary()function spits out the median of the replicate estimates. I don't know if that is customary (rather than the mean), but the output is not labelled as median, so I looked to documentation which is silent on the subject.