The tutorials can be found here:

• BigDataStatMeth to work on memory (Matrix and DelayedArray)
• BigDataStatMeth to work wiht HDF5 files

In order to reproduce the vignette follow the instructions described in the next sections

Our package needs to be installed from source code. In such cases, a collection of software (e.g. C, C++, Fortran, ...) are required, mainly for Windows users. These programs can be installed using Rtools.

Then, some required packages must be installed:

# Install BiocManager (if not previously installed)
install.packages("BiocManager") 

# Install required packages
BiocManager::install(c("Matrix", "RcppEigen", "RSpectra",
                       "beachmat", "DelayedArray",
                       "HDF5Array", "rhdf5"))

After that, BigDataStatMeth is installed from our GitHub repository:

# Install devtools and load library (if not previously installed)
install.packages("devtools") 
library(devtools)

# Install BigDataStatMeth 
install_github("isglobal-brge/BigDataStatMeth")

Download and execute this file.

Let us imaging we are interested fitting a linear model:

The ordinary least square (OLS) estimate of is

were is the QR decomposition of

To illustrate, let us consider the "mtcars" example, and run this regression:

data(mtcars)
lm(mpg ~ wt + cyl, data=mtcars)

Remeber that

Y <- matrix(mtcars$mpg)
X <- model.matrix(~ wt + cyl, data=mtcars)

Tasks:

• Use functions in R to estimate model parameters using OLS.
• Use functions in BigDataStatMeth to estimate model parameters using OLS with in memory data.
• Use functions in BigDataStatMeth to estimate model parameters using OLS with a HDF5 file.

NOTE: use bdpseudoinv() function instead of bdInvCholesky() since the matrix is not positive definite