variani

library(tidyverse)

# read selected columns: specify columns through 'col_types'
readr::read_csv("iris.csv", col_types = cols_only( Species = col_factor(c("setosa", "versicolor", "virginica"))))

variani

# Sparse Matrices in R: https://slowkow.com/notes/sparse-matrix/

library(igraph)
library(Matrix)
library(tidyverse)

#-------------------------------
# Adjacency matrix using Matrix
#-------------------------------

variani

#--------------------------------
# file operations with patterns
#--------------------------------

# list/delete files output/{chr}.out such that {chr} = 1-10 (other files with {chr} = 11-22 are untouched)
 ls output/{1..10}.* 


#--------------------
# misc. caveats

variani

# ------------------
# from bgen to bed
# ------------------
plink2 --bgen myfile.bgen \
  # filter by chr or by snps names
  --chr 1 --extract {params.snps} \ 
  --make-bed --out myoutput
# 1. plink will create temporary/intermediate files such as *-temporary.{pgen,psam,pvar}
# 2. sample ids in .fam will be `ID1_ID1`

variani

# print NAs as blank: https://stackoverflow.com/a/44348232

variani

# Sorting keys before look up speeds up queries.
# data.table outperforms over base R and tidyverse
#
# Copied from: https://appsilon.com/fast-data-lookups-in-r-dplyr-vs-data-table/
#
# Benchmarks:
#Unit: microseconds
#              expr       min        lq        mean    median        uq
#    look_up_base() 50762.103 53147.945 105500.0550 60221.954 62802.136
#   look_up_dplyr() 47527.238 49218.827  99142.9265 53923.058 58269.192

variani

library(magrittr)

data(mtcars)

lm(mpg ~ am, mtcars) %>% summary %$% coefficients
#             Estimate Std. Error   t value     Pr(>|t|)
#(Intercept) 17.147368   1.124603 15.247492 1.133983e-15
#am           7.244939   1.764422  4.106127 2.850207e-04

z <- 4.106127

variani

drop <- function(X)
{
  qr <- qr(X)
  r <- qr$rank # matrix rank

  cols_keep <- qr$pivot[seq(r)] # first r indices in `qr$pivot`
  #cols_drop <- qr$pivot[-seq(r)] # last (ncol(X) - r) indices in `qr$pivot`

  X[, cols_keep, drop = FALSE] # use drop = FALSE in single-column case
}

variani

# References
# - http://www.sthda.com/english/articles/31-principal-component-methods-in-r-practical-guide/118-principal-component-analysis-in-r-prcomp-vs-princomp/

## inc
library(tidyverse)
library(pls)

library(ggplot2)
library(cowplot)
theme_set(theme_cowplot())

variani

## examples to read tables by data.table::fread

# read selected columns
fread(my_file, select = c("ID", "FID"))

# read selected columns & specify colunm classes
fread(my_file, select = c("ID", "FID"), colClasses = c(IID = "character"))