stap_glm.Rd
Generalized linear modeling with spatial temporal aggregated predictors using prior distributions for the coefficients, intercept, spatialtemporal scales, and auxiliary parameters.
stap_glm(formula, family = gaussian(), subject_data = NULL, distance_data = NULL, time_data = NULL, subject_ID = NULL, max_distance = NULL, max_time = NULL, weights, offset = NULL, model = TRUE, y = TRUE, contrasts = NULL, ..., prior = normal(), prior_intercept = normal(), prior_stap = normal(), prior_theta = log_normal(location = 1L, scale = 1L), prior_aux = exponential(), adapt_delta = NULL) stap_lm(formula, subject_data = NULL, distance_data = NULL, time_data = NULL, subject_ID = NULL, max_distance = NULL, max_time = NULL, weights, offset = NULL, model = TRUE, y = TRUE, contrasts = NULL, ..., prior = normal(), prior_intercept = normal(), prior_stap = normal(), prior_theta = log_normal(location = 1L, scale = 1L), prior_aux = exponential(), adapt_delta = NULL)
formula  Same as for 


family  Same as 

subject_data  a data.frame that contains data specific to the subject or subjects on whom the outcome is measured. Must contain one column that has the subject_ID on which to join the distance and time_data 

distance_data  a (minimum) three column data.frame that contains (1) an id_key (2) The sap/tap/stap features and (3) the distances between subject with a given id and the built environment feature in column (2), the distance column must be the only column of type "double" and the sap/tap/stap features must be specified in the dataframe exactly as they are in the formula. 

time_data  same as distance_data except with time that the subject has been exposed to the built environment feature, instead of distance 

subject_ID  name of column(s) to join on between subject_data and bef_data 

max_distance  the inclusion distance; upper bound for all elements of dists_crs 

max_time  inclusion time; upper bound for all elements of times_crs 

offset, weights  Same as 

model  logical denoting whether or not to return the fixed covariates model frame object in the fitted object 

y  In 

contrasts  Same as 

...  Further arguments passed to the function in the rstap
to specify 

prior  The prior distribution for the regression coefficients.
See the priors help page for details on the families and
how to specify the arguments for all of the functions in the table above.
To omit a prior i.e., to use a flat (improper) uniform prior
Note: If 

prior_intercept  The prior distribution for the intercept.
Note: The prior distribution for the intercept is set so it
applies to the value when all predictors are centered. If you prefer
to specify a prior on the intercept without the predictors being
autocentered, then you have to omit the intercept from the


prior_stap  prior for spatialtemporal aggregated predictors. Note that prior is set on the standardized latent covariates. 

prior_theta  prior for the spatialtemporal aggregated predictors' scale and shape if the weibull weight function is selected. Can either be a single prior or a prior nested within a list of lists for separate BEFs/spacetime components. 

prior_aux  The prior distribution for the "auxiliary" parameter (if
applicable). The "auxiliary" parameter refers to a different parameter
depending on the


adapt_delta  See the adapt_delta help page for details. 
A stapreg object is returned
for stap_glm
.
A stapfit object (or a slightly modified
stapfit object) is returned if stan_glm.fit
is called directly.
The stap_glm
function is similar in syntax to
stan_glm
except instead of performing full bayesian
inference for a generalized linear model stap_glm incorporates spatialtemporal covariates
Gelman, A. and Hill, J. (2007). Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press, Cambridge, UK.
Muth, C., Oravecz, Z., and Gabry, J. (2018) Userfriendly Bayesian regression modeling: A tutorial with rstanarm and shinystan. The Quantitative Methods for Psychology. 14(2), 99119. https://www.tqmp.org/RegularArticles/vol142/p099/p099.pdf
stapregmethods
and
glm
.
The various vignettes for stap_glm
at
https://biostatistics4socialimpact.github.io/rstap/articles and the preprint article.
fit_glm < stap_glm(formula = y ~ sex + sap(Fast_Food), subject_data = homog_subject_data[1:100,], # for speed of example only distance_data = homog_distance_data, family = gaussian(link = 'identity'), subject_ID = 'subj_id', prior = normal(location = 0, scale = 5, autoscale = FALSE), prior_intercept = normal(location = 25, scale = 5, autoscale = FALSE), prior_stap = normal(location = 0, scale = 3, autoscale = FALSE), prior_theta = log_normal(location = 1, scale = 1), prior_aux = cauchy(location = 0,scale = 5), max_distance = max(homog_distance_data$Distance), chains = 1, iter = 300, # for speed of example only refresh = 1, verbose = FALSE)#> Chain 1: #> Chain 1: Gradient evaluation took 0.00039 seconds #> Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 3.9 seconds. #> Chain 1: Adjust your expectations accordingly! #> Chain 1: #> Chain 1: #> Chain 1: #> Chain 1: Elapsed Time: 0.811819 seconds (Warmup) #> Chain 1: 0.795173 seconds (Sampling) #> Chain 1: 1.60699 seconds (Total) #> Chain 1:#> Warning: Bulk Effective Samples Size (ESS) is too low, indicating posterior means and medians may be unreliable. #> Running the chains for more iterations may help. See #> http://mcstan.org/misc/warnings.html#bulkess#> Warning: Tail Effective Samples Size (ESS) is too low, indicating posterior variances and tail quantiles may be unreliable. #> Running the chains for more iterations may help. See #> http://mcstan.org/misc/warnings.html#tailess