find.climits {CTFS} | R Documentation |
Computes the confidence intervals for the number of surviving trees. This can be used to compute confidence intervals for rates of mortality and recruitment. The intervals are based on the beta distribution.
find.climits(N, S, alpha = 0.05, kind = "upper")
N |
number of individuals from a census |
S |
number of survivors |
alpha |
type I error |
kind |
type of confidence limit |
This function is called from mortality.calculation
and
recruitment.calculation
. On its own it computes the number
of surviving trees for one confidence interval of a given kind
for given values of N. For mortality, N should be
N0 the number of individuals alive in the first census. For
recruitment, N should be N1 the number of individuals
alive in the second census. For both mortality and recruitment,
S is the number of individuals that were alive in N0 AND
in N1. Since the confidence intervals provided are limits are
survivors, the lower confidence interval for survivorship provides the
upper confidence interval for mortality rate (high survivorship = lower
mortality) and upper confidence interval for recruitment rate (high
survivorship = lower number of recruits). And the upper confidence
interval for survivorship provides the lower confidence for mortality
(low survivorship = higher mortality) and the lower confidence interval
for recruitment rate (low survivorship = higher number of recruits).
The confidence limits for survivorship are computed using a beta distribution.
Returns the confidence limits in number of surviving trees:
S |
numeric, the best estimate of the number of surviving trees for the given kind of limit. |
Rick Condit and Pamela Hall
CTFS.mortality
, CTFS.recruitment
## Not run: # N0 = 100, Dead = 5, Recruits = 10, time = 5 yrs, therefore: # S = N0 - Dead = 100 - 5 = 95 # N1 = N0 - Dead + Recruits = 100 - 5 + 10 = 105 # find CI limits for mortality SM.upper <- find.climits(100,95,kind="upper") SM.lower <- find.climits(100,95,kind="lower") mort.rate.upper <- ( log(N0) - log(SM.lower) ) / 5 mort.rate.lower <- ( log(N0) - log(SM.upper) ) / 5 # find CI limits for recruitment SR.upper <- find.climits(105,95,kind="upper") SR.lower <- find.climits(105,95,kind="lower") rec.rate.upper <- ( log(N1) - log(SR.lower) ) / 5 rec.rate.lower <- ( log(N1) - log(SR.upper) ) / 5 ## End(Not run)