Get RMT threshold for a correlation matrix

Get RMT threshold for a correlation matrix roughly

RMT_threshold(
  occor.r,
  out_dir,
  min_threshold = 0.5,
  max_threshold = 0.8,
  step = 0.02,
  gif = FALSE,
  verbose = FALSE
)

rmt(occor.r, min_threshold = 0.5, max_threshold = 0.85, step = 0.01)

Arguments

occor.r: a corr object or a correlation matrix
out_dir: output dir
min_threshold: min_threshold
max_threshold: max_threshold
step: step
gif: render a .gif file?
verbose: verbose

Value

a r-threshold

recommend threshold

References

J. Zhou, Y. Deng, FALSE. Luo, Z. He, Q. Tu, X. Zhi, (2010) Functional Molecular Ecological Networks, doi:10.1128/mBio.00169-10. https://matstat.org/content_en/RMT/RMThreshold_Intro.pdf

Examples

# \donttest{
data(otutab, package = "pcutils")
t(otutab) -> totu
c_net_calculate(totu) -> corr
rmt(corr)
#> Calculating: 0.5
#> Calculating: 0.6
#> Calculating: 0.65
#> Calculating: 0.7
#> Calculating: 0.75
#> Calculating: 0.8
#> Calculating: 0.85

#> We recommend r-threshold: 0.69, you can calculate again in a smaller region

#> [1] 0.69
# recommend: 0.69
c_net_build(corr, r_threshold = 0.69) -> co_net_rmt
#> Have not do p-value adjustment! use the p.value to build network.
# }