albanesere / PandemicLP

Models for epidemic/pandemic data prediction using a generalized logistic curve for the counts average

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Installing PandemicLP

Installing binaries

If your system is Windows or MacOS, you can install the binaries. Make sure your R version is updated to 4.0.2. If your R version is lower, then you are recommended to install the package from source, as instructed below. For installing the binaries, run the appropriate code for your version.

# Windows
install.packages("http://github.com/CovidLP/PandemicLP/raw/master/binaries/PandemicLP_0.1.0.zip",repos=NULL)

# MacOS
install.packages("http://github.com/CovidLP/PandemicLP/raw/master/binaries/PandemicLP_0.1.0.tgz",repos=NULL)

Note that the vignette will be readily available after this command is done running through the command vignette("PandemicLP").

Installing from source

First, make sure you have the devtools package installed by typing

install.packages("devtools")

Then run code

devtools::install_github("CovidLP/PandemicLP", build_vignettes = TRUE)

The option build_vignettes = TRUE will make the installation take longer, but it will make the example vignette available for viewing.

V8 dependency upon building from source

Current version of rstan depends on the V8 JavaScript libraries. In order to install rstan, you might require to install it in your Unix-based system. On Windows, you might need to install R package V8 from the CRAN binaries by running install.packages("V8").

If installation fails on Mac OS

Installation on Mac requires 3.6.2, so make sure it is up to date.

If you are using the Catalina version of the Mac Operating System, go to here and follow the same instructions for installing rstan from source.

If installation fails on Windows

In RStudio (preferably) or otherwise in R, execute once

pkgbuild::has_build_tools(debug = TRUE)

to check your C++ toolchain using the pkgbuild package that gets installed when you install RStan.

Then,

  1. If you are using Rstudio, a pop-up will appear asking if you want to install Rtools, if you do not already have it. Click Yes and wait about a minute until the installation is finished.
  2. If you are not using RStudio, a message will appear in the R console telling you to install Rtools.

In some cases, Windows requires this block of code to be run in R for the installation to be successful. However, do not do this unless the installation has failed.

dotR <- file.path(Sys.getenv("HOME"), ".R")
if (!file.exists(dotR)) dir.create(dotR)
M <- file.path(dotR, ifelse(.Platform$OS.type == "windows", "Makevars.win", "Makevars"))
if (!file.exists(M)) file.create(M)
cat(if( grepl("^darwin", R.version$os)) "\nCXX14FLAGS=-O3 -march=native -mtune=native -arch x86_64 -ftemplate-depth-256" else 
    if (.Platform$OS.type == "windows") "\nCXX14FLAGS=-O3 -mtune=native -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2" else
    "CXX14FLAGS = -fPIC",
    file = M, sep = "\n", append = TRUE)

If installation still fails

If the steps above were tried for your OS and it still didn't work, please e-mail us at covidlp.team@gmail.com

About

Models for epidemic/pandemic data prediction using a generalized logistic curve for the counts average


Languages

Language:R 74.7%Language:C++ 24.4%Language:Stan 0.9%