===========================================================
.___ __ __
_________________ __ __ __| _/|__|/ |_
/ ___\_` __ \__ \ | | \/ __ | | \\_ __\
/ /_/ > | \// __ \| | / /_/ | | || |
\___ /|__| (____ /____/\____ | |__||__|
/_____/ \/ \/
grep rough audit - static analysis tool
v2.8 written by @Wireghoul
=================================[justanotherhacker.com]===
##############################################
r-cran-surveillance-1.18.0/vignettes/surveillance.Rnw-163-for example with a Salmonella serovar. The model is a Hidden Markov
r-cran-surveillance-1.18.0/vignettes/surveillance.Rnw:164:Model (HMM) where a binary state $X_t, t=1,\ldots,n$, denotes whether
r-cran-surveillance-1.18.0/vignettes/surveillance.Rnw-165-there was an outbreak and $Y_t$ is the number of observed
##############################################
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw-296- $\left(y_{t_0-7},\ldots,y_{t_0-1}\right)$. The expected value is the mean of the baseline. The method is based on a statistic called $C_{t_0}$ defined as
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw:297:$C_{t_0}= \frac{(y_{t_0}-\bar{y}_{t_0})}{s_{t_0}}$, where $$\bar{y}_{t_0}= \frac{1}{7} \cdot\sum_{i=t_0-7}^{t_0-1} y_i \textnormal{ and } s_{t_0}^2= \frac{1}{7-1} \cdot\sum_{i=t_0-7}^{t_0-1} \left(y_i - \bar{y}_{t_0}\right)^2.$$
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw-298-Under the null hypothesis of no outbreak, it is assumed that $C_{t_0} \stackrel{H_0}{\sim} {N}(0,1)$. The upperbound $U_{t_0}$ is found by assuming that $y_t$ is normal,
##############################################
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw-380-In both cases the steps of the algorithm are the same. In a first step, an overdispersed Poisson generalized linear model with log link is fitted to the reference data $\bm{y}_{t_0} \subseteq \left\{
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw:381:y_t\>;\> t\leq t_0\right\}$, where $\E(y_t)=\mu_t$ with $\log \mu_t = \alpha + \beta t$ and $\Var(y_t)=\phi\cdot\mu_t$ and where $\phi\geq1$ is ensured.
r-cran-surveillance-1.18.0/vignettes/monitoringCounts.Rnw-382-
##############################################
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-85-
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw:86:Consider the situation, where a time series of counts is collected for surveillance purpose. In each interval, usually one week, the number of cases of the interesting disease in an area (country, district) is counted. The resulting time series is denoted by $\{y_t\>;t=1,\ldots,n\}$. Usually the data are collected on line, so that the time point $n$ is the actual time point. Our aim is to decide with the aid of a statistic for each time point $n$ if there is an outbreak at this or any former time point. If an outbreak is detected, the algorithm gives an alarm. Observed time series of counts are saved in a \verb+disProg+ object, a list containing the time series of counts, the number of weeks and a state chain.
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-87-The state is 1, if e.g. the Robert Koch Institut declares the week to be part of an outbreak and 0 otherwise ~\citep{survstat}. By using the state chain the quality of the surveillance algorithm can be tested.
##############################################
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-132-
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw:133:Now $N=\inf \{n \geq 1 : GLR(n) \geq c_{\gamma} \}$ is the first time point where the GLR-statistic is above a threshold $c_{\gamma}$. For this time point $N$ an alarm is given. If the parameter $\kappa$ and hence $\theta=\kappa$ is known, the maximisation over $\theta$ can be omitted.
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-134-
##############################################
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-363-\item \verb+Mtilde+ and \verb+M+ \\
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw:364:These parameters are necessary for the so called ''window-limited'' GLR scheme. Here the maximation is not performed for all $1 \leq k \leq n$ but instead only for a window $k \in \{n-M,...,n-\tilde{M}+1 \}$ of values. Note that $1 \leq \tilde{M} \leq M$, where the minimum delay $\tilde{M}$ is the minimal required sample size to obtain a sufficient estimate of $\theta_1=(\mu_0,\kappa)$ ~\citep{hoehle.paul2008}. The advantage of using a window of values instead of all values is the faster computation, but in the setup with intercept-charts and $\theta_1=\kappa$ this doesn't bother much and $\tilde{M}=1$ is sufficient.
r-cran-surveillance-1.18.0/vignettes/glrnb.Rnw-365-
##############################################
r-cran-surveillance-1.18.0/R/algo_farrington.R-315- ######################################################################
r-cran-surveillance-1.18.0/R/algo_farrington.R:316: defaultControl <- eval(formals()$control)
r-cran-surveillance-1.18.0/R/algo_farrington.R-317- control <- modifyList(defaultControl, control, keep.null = TRUE)
##############################################
r-cran-surveillance-1.18.0/R/algo_glrnb.R-266- message(paste0("glrnb: Fitting glm.nb model with alpha=",control$alpha))
r-cran-surveillance-1.18.0/R/algo_glrnb.R:267: m <- eval(substitute(glm(form,family=negative.binomial(theta=1/control$alpha),data=data),list(form=as.formula(formula))))
r-cran-surveillance-1.18.0/R/algo_glrnb.R-268- }
##############################################
r-cran-surveillance-1.18.0/R/stsplot_space.R-70- if (!missing(colorkey) && is.list(colorkey))
r-cran-surveillance-1.18.0/R/stsplot_space.R:71: colorkey <- modifyList(eval(formals()$colorkey), colorkey)
r-cran-surveillance-1.18.0/R/stsplot_space.R-72-
##############################################
r-cran-surveillance-1.18.0/R/hhh4_methods.R-348- ## object$control$start is not necessarily a complete list:
r-cran-surveillance-1.18.0/R/hhh4_methods.R:349: template <- eval(formals(hhh4)$control$start)
r-cran-surveillance-1.18.0/R/hhh4_methods.R-350- template[] <- object$control$start[names(template)]
##############################################
r-cran-surveillance-1.18.0/R/epidataCS_plot.R-206- } else { # calls sp:::subset.Spatial
r-cran-surveillance-1.18.0/R/epidataCS_plot.R:207: eval(substitute(base::subset(x$events, subset=.subset),
r-cran-surveillance-1.18.0/R/epidataCS_plot.R-208- list(.subset=substitute(subset))))
##############################################
r-cran-surveillance-1.18.0/R/twinSIR.R-251- ## Verify that 'data' inherits from "epidata"
r-cran-surveillance-1.18.0/R/twinSIR.R:252: data <- eval(cl$data, parent.frame())
r-cran-surveillance-1.18.0/R/twinSIR.R-253- if (!inherits(data, "epidata")) {
##############################################
r-cran-surveillance-1.18.0/R/epidataCS_methods.R-144- e <- substitute(subset)
r-cran-surveillance-1.18.0/R/epidataCS_methods.R:145: r <- eval(e, x$events@data, parent.frame()) # HERE IS A MOD
r-cran-surveillance-1.18.0/R/epidataCS_methods.R-146- if (!is.logical(r)) stop("'subset' must evaluate to logical")
##############################################
r-cran-surveillance-1.18.0/R/hhh4.R-172- ## arguments in 'control' override any corresponding default arguments
r-cran-surveillance-1.18.0/R/hhh4.R:173: defaultControl <- eval(formals(hhh4)$control)
r-cran-surveillance-1.18.0/R/hhh4.R-174- environment(defaultControl$ar$f) <- environment(defaultControl$ne$f) <-
##############################################
r-cran-surveillance-1.18.0/R/twinstim_epitest.R-21- method <- match.arg(method, choices = c("LRT", "simulate", "time", "space"))
r-cran-surveillance-1.18.0/R/twinstim_epitest.R:22: # eval(formals(permute.epidataCS)$what)
r-cran-surveillance-1.18.0/R/twinstim_epitest.R-23- if (model$npars["q"] == 0L) {
##############################################
r-cran-surveillance-1.18.0/R/farringtonFlexible.R-102-
r-cran-surveillance-1.18.0/R/farringtonFlexible.R:103: defaultControl <- eval(formals()$control)
r-cran-surveillance-1.18.0/R/farringtonFlexible.R-104- control <- modifyList(defaultControl, control, keep.null = TRUE)
##############################################
r-cran-surveillance-1.18.0/R/sts_animate.R-39- if (is.list(timeplot)) {
r-cran-surveillance-1.18.0/R/sts_animate.R:40: timeplot <- modifyList(eval(formals()$timeplot), timeplot)
r-cran-surveillance-1.18.0/R/sts_animate.R-41- timeplot_height <- timeplot$height
##############################################
r-cran-surveillance-1.18.0/R/twinstim_methods.R-317- which <- match.arg(which, choices =
r-cran-surveillance-1.18.0/R/twinstim_methods.R:318: c(eval(formals(intensityplot.twinstim)$which),
r-cran-surveillance-1.18.0/R/twinstim_methods.R:319: eval(formals(iafplot)$which)))
r-cran-surveillance-1.18.0/R/twinstim_methods.R:320: FUN <- if (which %in% eval(formals(intensityplot.twinstim)$which))
r-cran-surveillance-1.18.0/R/twinstim_methods.R-321- "intensityplot" else "iafplot"
##############################################
r-cran-surveillance-1.18.0/R/twinstim_simulation.R-133-
r-cran-surveillance-1.18.0/R/twinstim_simulation.R:134: if (is.null(t0)) t0 <- eval(formals()$t0)
r-cran-surveillance-1.18.0/R/twinstim_simulation.R:135: if (is.null(T)) T <- eval(formals()$T)
r-cran-surveillance-1.18.0/R/twinstim_simulation.R-136- if (!isScalar(t0) || !isScalar(T)) {
##############################################
r-cran-surveillance-1.18.0/R/twinstim_simulation.R-1004- modelenv$nTypes)
r-cran-surveillance-1.18.0/R/twinstim_simulation.R:1005: hGrid <- sum(exp(.beta0)) * eval(modelenv$hGridExpr, envir = modelenv)
r-cran-surveillance-1.18.0/R/twinstim_simulation.R-1006- blockstartstop <- modelenv$histIntervals[
##############################################
r-cran-surveillance-1.18.0/inst/doc/surveillance.Rnw-163-for example with a Salmonella serovar. The model is a Hidden Markov
r-cran-surveillance-1.18.0/inst/doc/surveillance.Rnw:164:Model (HMM) where a binary state $X_t, t=1,\ldots,n$, denotes whether
r-cran-surveillance-1.18.0/inst/doc/surveillance.Rnw-165-there was an outbreak and $Y_t$ is the number of observed
##############################################
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw-296- $\left(y_{t_0-7},\ldots,y_{t_0-1}\right)$. The expected value is the mean of the baseline. The method is based on a statistic called $C_{t_0}$ defined as
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw:297:$C_{t_0}= \frac{(y_{t_0}-\bar{y}_{t_0})}{s_{t_0}}$, where $$\bar{y}_{t_0}= \frac{1}{7} \cdot\sum_{i=t_0-7}^{t_0-1} y_i \textnormal{ and } s_{t_0}^2= \frac{1}{7-1} \cdot\sum_{i=t_0-7}^{t_0-1} \left(y_i - \bar{y}_{t_0}\right)^2.$$
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw-298-Under the null hypothesis of no outbreak, it is assumed that $C_{t_0} \stackrel{H_0}{\sim} {N}(0,1)$. The upperbound $U_{t_0}$ is found by assuming that $y_t$ is normal,
##############################################
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw-380-In both cases the steps of the algorithm are the same. In a first step, an overdispersed Poisson generalized linear model with log link is fitted to the reference data $\bm{y}_{t_0} \subseteq \left\{
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw:381:y_t\>;\> t\leq t_0\right\}$, where $\E(y_t)=\mu_t$ with $\log \mu_t = \alpha + \beta t$ and $\Var(y_t)=\phi\cdot\mu_t$ and where $\phi\geq1$ is ensured.
r-cran-surveillance-1.18.0/inst/doc/monitoringCounts.Rnw-382-
##############################################
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-85-
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw:86:Consider the situation, where a time series of counts is collected for surveillance purpose. In each interval, usually one week, the number of cases of the interesting disease in an area (country, district) is counted. The resulting time series is denoted by $\{y_t\>;t=1,\ldots,n\}$. Usually the data are collected on line, so that the time point $n$ is the actual time point. Our aim is to decide with the aid of a statistic for each time point $n$ if there is an outbreak at this or any former time point. If an outbreak is detected, the algorithm gives an alarm. Observed time series of counts are saved in a \verb+disProg+ object, a list containing the time series of counts, the number of weeks and a state chain.
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-87-The state is 1, if e.g. the Robert Koch Institut declares the week to be part of an outbreak and 0 otherwise ~\citep{survstat}. By using the state chain the quality of the surveillance algorithm can be tested.
##############################################
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-132-
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw:133:Now $N=\inf \{n \geq 1 : GLR(n) \geq c_{\gamma} \}$ is the first time point where the GLR-statistic is above a threshold $c_{\gamma}$. For this time point $N$ an alarm is given. If the parameter $\kappa$ and hence $\theta=\kappa$ is known, the maximisation over $\theta$ can be omitted.
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-134-
##############################################
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-363-\item \verb+Mtilde+ and \verb+M+ \\
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw:364:These parameters are necessary for the so called ''window-limited'' GLR scheme. Here the maximation is not performed for all $1 \leq k \leq n$ but instead only for a window $k \in \{n-M,...,n-\tilde{M}+1 \}$ of values. Note that $1 \leq \tilde{M} \leq M$, where the minimum delay $\tilde{M}$ is the minimal required sample size to obtain a sufficient estimate of $\theta_1=(\mu_0,\kappa)$ ~\citep{hoehle.paul2008}. The advantage of using a window of values instead of all values is the faster computation, but in the setup with intercept-charts and $\theta_1=\kappa$ this doesn't bother much and $\tilde{M}=1$ is sufficient.
r-cran-surveillance-1.18.0/inst/doc/glrnb.Rnw-365-
##############################################
r-cran-surveillance-1.18.0/debian/tests/run-unit-test-5-if [ "$AUTOPKGTEST_TMP" = "" ] ; then
r-cran-surveillance-1.18.0/debian/tests/run-unit-test:6: AUTOPKGTEST_TMP=`mktemp -d /tmp/${pkg}-test.XXXXXX`
r-cran-surveillance-1.18.0/debian/tests/run-unit-test-7-fi