assemble.data.lags: Assemble human case data and environmental lags

In khelmsmith/flm_NE_WNV: Functional Linear Modeling For Vector-borne Disease

Assemble human case data and environmental lags

Description

assemble.data.lags() expects inputs to follow certain conventions (see below).

Usage

assemble.data.lags(
  pop,
  cases,
  weather,
  spi,
  spei,
  target.date,
  start.year,
  in.seed = NULL,
  lag.lengths = c(12, 18, 24, 30, 36)
)

Arguments

pop	County populations, a data frame with 5 variables County , fips (5 characters), year , pop100K , density .
cases	data on annual numbers of human cases in each county. A data.frame with 3 variables, County , year , and cases .
weather	monthly temperature and precipitation data for for each county. A data.frame with County , fips (5 characters), year , month (integer), tmean , and ppt .
spi	monthly values of the Standardized Precipitation Index for each county. County , fips (5 characters), year , month (integer), spi .
spei	monthly values of the Standardized Precipitation and Evapotranspiration Index for each county. County , fips (5 characters), year , month (integer), spei .
target.date	The last date to include for calculation of lags, a character string with ISO XXX format (yyyy-mm-dd).
start.year	The first year to include in the training data. Should be coercible to integer.
in.seed	If not NULL, the starting number for the random number generator. This makes the results repeatable. If NULL, treats cases as actual data
lag.lengths	the number of months to go backwards when creating lag matrices. Numeric vector.

Details

Case data are trimmed to between

start.year

and

target.year

. predictor variables are trimmed to between January of

start.year

and the date specified in

target.year

.

The

County

variable can be any character string unique to each geographic locale, as long as the format is identical across all input data frames. Simulated data (in.seed not null) is predictions of a model that was trained on actual numbers of cases as recorded in CDC's Arbonet database. It excludes Arthur County, because no cases have been recorded there to date, and we had to exclude it from our modeling to get it to work. Weather data included in package for Nebraska comes from National Centers for Environmental Information, National Climatic Data Center (ftp://ftp.ncdc.noaa.gov/pub/data/cirs/climdiv/). SPEI and SPI from Westwide Drought Tracker netcdf files. See Abatzoglou, J. T., McEvoy, D. J., & Redmond, K. T. (2017). The West Wide Drought Tracker: Drought monitoring at fine spatial scales. Bulletin of the American Meteorological Society, 98(9), 1815–1820. https://doi.org/10.1175/BAMS-D-16-0193.1.

Value

a list of two data frames, one with training data and one with the target year data.

khelmsmith/flm_NE_WNV documentation built on June 24, 2022, 11:13 p.m.