, Caitlin Campbell
[ctb] , Daijiang Li
[ctb] | Title: | Estimate Phenological Metrics using Presence-Only Data |
|---|---|
| Description: | Generates Weibull-parameterized estimates of phenology for any percentile of a distribution using the framework established in Cooke (1979) <doi:10.1093/biomet/66.2.367>. Extensive testing against other estimators suggest the weib_percentile() function is especially useful in generating more accurate and less biased estimates of onset and offset (Belitz et al. 2020 <doi.org:10.1111/2041-210X.13448>. Non-parametric bootstrapping can be used to generate confidence intervals around those estimates, although this is computationally expensive. Additionally, this package offers an easy way to perform non-parametric bootstrapping to generate confidence intervals for quantile estimates, mean estimates, or any statistical function of interest. |
| Authors: | Michael Belitz [aut, cre]
, Caitlin Campbell
[ctb] , Daijiang Li
[ctb] |
| Maintainer: | Michael Belitz <[email protected]> |
| License: | CC0 |
| Version: | 0.1.2 |
| Built: | 2024-11-02 03:42:36 UTC |
| Source: | https://github.com/cran/phenesse |
Calculate confidence intervals using bootstrap of any statistical function of interest.
estimate_ci( observations, .f, n_boots, parallelize = "no", ncpus = getOption("boot.ncpus", 1L), cl = NULL, type = "perc", conf = 0.95 )estimate_ci( observations, .f, n_boots, parallelize = "no", ncpus = getOption("boot.ncpus", 1L), cl = NULL, type = "perc", conf = 0.95 )
observations |
is a vector of dates/time of observations given as numeric values |
.f |
function to use |
n_boots |
is the number of bootstraps you want to run to create the CIs |
parallelize |
The type of parallel operation to be used (if any). If missing, the default is that no parallelization will occur. Parallelization options are "multicore" and "snow" |
ncpus |
An integer that represents the number of processes to be used in parallel operation. |
cl |
An optional parallel or snow cluster for use if parallel = "snow". If not supplied, a cluster on the local machine is created for the duration of the boot call. |
type |
A vector of character strings representing the type of intervals required to calculate the CI. Defaults to "perc". See ??boot.ci for more information. |
conf |
The confidence level wanted. Defaults to 95% CI. |
A data frame with estimate, and the lower and upper points of its confidence interval
Example data downloaded from iNaturalist.org for the Washington, DC area using the bounding box bounds = c(38, -77, 39, -76). Data was downloaded on 10/30/2019 for four species: Speyeria cybele, Danaus plexippus, Rudbeckia hirta, and Asclepias syriaca. These data were not scored to mark phenology, so all life stages/reproductive stages are included in the download.The download only includes 2019 observations and the doy (day of year)column was added post data download by MW Belitz using the lubridate package.
data(inat_examples)data(inat_examples)
A data frame with 252 rows and 6 variables:
binomial of species
latitude where observations occurred
longitude where observations occurred
common name related to species
original date listed of observation
day of year when the observation occurred, variable created by MW Belitz using the package lubridate
data(inat_examples) ## Not run: View(inat_examples) ## End(Not run)data(inat_examples) ## Not run: View(inat_examples) ## End(Not run)
mean_ciFunction estimates CIs using nonparametric bootstrapping around a
mean estimate.
mean_ci(observations, bootstraps = 1e+05, conf = 0.95, type = "perc")mean_ci(observations, bootstraps = 1e+05, conf = 0.95, type = "perc")
observations |
A vector of observations given as numeric values |
bootstraps |
The number of bootstraps you want to run to create the CIs, defaults to 100000 |
conf |
The confidence level wanted. Defaults to 95% CI. |
type |
A vector of character strings representing the type of intervals required to calculate the CI. Defaults to "bca". See ??boot.ci for more information. |
The estimated CIs around a mean estimate.
mean_ci: Estimates CIs around a mean percentile estimate using
non-parametric bootstrapping from the boot package
# Estimate when the mean observation of Rudbeckia hirta for the year 2019 up # to October data(inat_examples) r_hirta <- subset(inat_examples, scientific_name == "Rudbeckia hirta") mean_ci(observations = r_hirta$doy , bootstraps = 100) # note low number of bootstraps for quick processing speed# Estimate when the mean observation of Rudbeckia hirta for the year 2019 up # to October data(inat_examples) r_hirta <- subset(inat_examples, scientific_name == "Rudbeckia hirta") mean_ci(observations = r_hirta$doy , bootstraps = 100) # note low number of bootstraps for quick processing speed
quantile_ciEstimates CIs around a quantile percentile estimate using
non-parametric bootstrapping from the boot package
quantile_ci( observations, percentile, bootstraps = 1e+05, conf = 0.95, type = "perc" )quantile_ci( observations, percentile, bootstraps = 1e+05, conf = 0.95, type = "perc" )
observations |
A vector of observations given as numeric values |
percentile |
The percentile of interest |
bootstraps |
The number of bootstraps you want to run to create the CIs, defaults to 100000 |
conf |
The confidence level wanted. Defaults to 95% CI. |
type |
A vector of character strings representing the type of intervals required to calculate the CI. Defaults to "bca". See ??boot.ci for more information. |
The quantile estimate and confidence intervals.
# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 10 percent of individuals of the butterfly species # Speyeria cybele are in flight. data(inat_examples) s_cybele <- subset(inat_examples, scientific_name == "Speyeria cybele") quantile_ci(observations = s_cybele$doy, percentile = 0.1, bootstraps = 100) # note low number of bootstraps for quick processing speed# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 10 percent of individuals of the butterfly species # Speyeria cybele are in flight. data(inat_examples) s_cybele <- subset(inat_examples, scientific_name == "Speyeria cybele") quantile_ci(observations = s_cybele$doy, percentile = 0.1, bootstraps = 100) # note low number of bootstraps for quick processing speed
weib_percentile uses empirical bootstrapping to estimate a percentile
of the Weibull distribution, given random variables.
weib_percentile(observations, percentile, iterations = 500)weib_percentile(observations, percentile, iterations = 500)
observations |
is a vector of dates/time of observations given as integers |
percentile |
is the percentile of the cumulative distribution function of interest |
iterations |
is the number of iterations you want to use to bootstrap an estimate of bias of the original CDF. The bias is used to calculate a Weibull-corrected estimate of the percentile bound. |
The Weibull-corrected estimate of the percentile of interest.
# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 50 percent of individuals of the milkweed species # Asclepias syriaca have been observed. data(inat_examples) a_syriaca <- subset(inat_examples, scientific_name == "Asclepias syriaca") weib_percentile(a_syriaca$doy, percentile = 0.5, iterations = 500) # Estimate when 90 percent of individuals of the milkweed species A. syriaca # have been observed, using only 100 iterations for quicker processing. To # get a more stable result, more iterations should be used. weib_percentile(a_syriaca$doy, percentile = 0.9, iterations = 100)# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 50 percent of individuals of the milkweed species # Asclepias syriaca have been observed. data(inat_examples) a_syriaca <- subset(inat_examples, scientific_name == "Asclepias syriaca") weib_percentile(a_syriaca$doy, percentile = 0.5, iterations = 500) # Estimate when 90 percent of individuals of the milkweed species A. syriaca # have been observed, using only 100 iterations for quicker processing. To # get a more stable result, more iterations should be used. weib_percentile(a_syriaca$doy, percentile = 0.9, iterations = 100)
weib_percentile_ci uses non-parametric bootstrapping
from the boot package to estimate 95
weib_percentile_ci( observations, iterations, percentile, bootstraps, type = "perc", conf = 0.95, parallelize = "no", ncpus = getOption("boot.ncpus", 1L), cl = NULL )weib_percentile_ci( observations, iterations, percentile, bootstraps, type = "perc", conf = 0.95, parallelize = "no", ncpus = getOption("boot.ncpus", 1L), cl = NULL )
observations |
is a vector of dates/time of observations given as numeric values |
iterations |
is the number of iterations you want to run to create empirical bootstrapping to estimate bias of original CDF. The bias is used to calculate a bias corrected estimate of the percentile bound. |
percentile |
is the percentile of the cumulative distribution function of interest |
bootstraps |
is the number of bootstraps you want to run to create the CIs |
type |
A vector of character strings representing the type of intervals required to calculate the CI. Defaults to "perc". See ??boot.ci for more information. |
conf |
The confidence level wanted. Defaults to 95% CI. |
parallelize |
The type of parallel operation to be used (if any). If missing, the default is that no parallelization will occur. Parallelization options are "multicore" and "snow" |
ncpus |
An integer that represents the number of processes to be used in parallel operation. |
cl |
An optional parallel or snow cluster for use if parallel = "snow". If not supplied, a cluster on the local machine is created for the duration of the boot call. |
The Weibull-corrected estimate of the percentile of interest and CIs.
# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 50 percent of individuals of the butterfly species # Speyeria cybele are in flight. Note, only 10 iterations are beingg used # in this example to keep computation time low, but more iterations should # be used to get a more stable result. See vignette for parallelization # options. data(inat_examples) s_cybele <- subset(inat_examples, scientific_name == "Speyeria cybele") weib_percentile_ci(observations = s_cybele$doy, iterations = 10, percentile = 0.5, bootstraps = 100)# Gather sightings of iNaturalist observations for four species: # Danaus plexippus, Speyeria cybele, Rudbeckia hirta, and Asclepias syriaca # Estimate when the first 50 percent of individuals of the butterfly species # Speyeria cybele are in flight. Note, only 10 iterations are beingg used # in this example to keep computation time low, but more iterations should # be used to get a more stable result. See vignette for parallelization # options. data(inat_examples) s_cybele <- subset(inat_examples, scientific_name == "Speyeria cybele") weib_percentile_ci(observations = s_cybele$doy, iterations = 10, percentile = 0.5, bootstraps = 100)