SpatioTemporal (version 0.9.2)

mesa.data.raw: Example of raw data

Description

The raw data that was used to create the mesa.data and mesa.data.model structures.

Usage

data(mesa.data.raw)

encoding

latin1

source

Contains monitoring data from the MESA Air project, see Cohen et.al. (2009) for details.

Details

The data structure contains raw data from the MESA Air project. The example below describes how to create the mesa.data structure from raw data.

References

M. A. Cohen, S. D. Adar, R. W. Allen, E. Avol, C. L. Curl, T. Gould, D. Hardie, A. Ho, P. Kinney, T. V. Larson, P. D. Sampson, L. Sheppard, K. D. Stukovsky, S. S. Swan, L. S. Liu, J. D. Kaufman. (2009) Approach to Estimating Participant Pollutant Exposures in the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air). Environmental Science & Technology: 43(13), 4687-4693.

See Also

See mesa.data and create.data.model.

For other data functions, see mesa.data.model, create.data.matrix, construct.LUR.basis and default.LUR.list.

Examples

##load the data
data(mesa.data.raw)

##create the location data.frame
location <- mesa.data.raw$X[,c("ID","x","y","long","lat","type")]
##ensure that it's a data.frame
location <- as.data.frame(location)
##ensure that ID and ID are factors
location$ID <- factor(as.character(location$ID))
location$type <- factor(as.character(location$type))

##extract the covariate LUR matrix
LUR <- mesa.data.raw$X
##ensure that it's a data.frame
LUR <- as.data.frame(LUR)
##set rownames to the IDs and drop location information
rownames(LUR) <- as.character(LUR$ID)
LUR <- LUR[,!(names(LUR) %in% c("ID","x","y","long","lat","type"))]

##create matrices of observation times
T <- t(matrix(rownames(mesa.data.raw$obs),
              nrow=dim(mesa.data.raw$obs)[2],
              ncol=dim(mesa.data.raw$obs)[1],byrow=TRUE))
##...and locations
ID <- matrix(colnames(mesa.data.raw$obs), nrow=dim(mesa.data.raw$obs)[1],
             ncol=dim(mesa.data.raw$obs)[2],byrow=TRUE)
##let's study these matrices
print(T[1:5,1:5])
print(ID[1:5,1:5])

##combine with the observations
obs <- data.frame(obs=c(mesa.data.raw$obs), date=as.Date(T),
                  ID=c(ID))
##drop unmonitored locations
obs <- obs[!is.na(obs$obs),,drop=FALSE]
##sort the locations (strictly not needed)
obs <- obs[order(obs$date,obs$ID),,drop=FALSE]

##compute the smooth trends
trend <- calc.smooth.trends(obs=obs$obs, date=obs$date,
                            ID=obs$ID, n.basis = 2)$svd

##create a 3D-array for the spatio-temporal covariate
ST <- array(mesa.data.raw$lax.conc.1500, dim =
            c(dim(mesa.data.raw$lax.conc.1500),1))
##add names for time, location and which spatio-temporal covariate
dimnames(ST) <- list(rownames(mesa.data.raw$lax.conc),
                     colnames(mesa.data.raw$lax.conc),
                     "lax.conc.1500")

##combine to a mesa.data-object.
mesa.data.alt <- list(location=location, LUR=LUR, trend = trend,
                      obs=obs, SpatioTemp=ST)

##This should be the same as the data in mesa.data