Databases
StackOverflow2013
Postsdbo
POBootstrap variables correlation in R
Body
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POBootstrap variables correlation in R
    text
    copied!<p>My intention was to write several functions aimed at finding the overall similarity between two covariance matrices, either by multiplying them with random vectors and correlating the response vectors or by bootstrapping one of the matrices to obtain the correlation coefficient distribution that can serve for comparison. But in both cases I got erroneous results. The observed between-matrix correlation was high up to 0.93, but the distribution only ranged up to 0.2 the most. This is the function`s code:</p> <pre><code>resamplerSimAlt &lt;- function(mat1, mat2, numR, graph = FALSE) { statSim &lt;- numeric(numR) mat1vcv &lt;- cov(mat1) mat2vcvT &lt;- cov(mat2) ltM1 &lt;- mat1vcv[col(mat1vcv) &lt;= row(mat1vcv)] ltM2T &lt;- mat2vcvT[col(mat2vcvT) &lt;= row(mat2vcvT)] statObs &lt;- cor(ltM1, ltM2T) indice &lt;- c(1:length(mat2)) resamplesIndices &lt;- lapply(1:numR, function(i) sample(indice, replace = F)) for (i in 1:numR) { ss &lt;- mat2[sample(resamplesIndices[[i]])] ss &lt;- matrix(ss, nrow = dim(mat2)[[1]], ncol = dim(mat2)[[2]]) mat2ss &lt;- cov(ss) ltM2ss &lt;- mat2ss[col(mat2ss) &lt;= row(mat2ss)] statSim[i] &lt;- cor(ltM1, ltM2ss) } if (graph == TRUE) { plot(1, main = "resampled data density distribution", xlim = c(0, statObs+0.1), ylim = c(0,14)) points(density(statSim), type="l", lwd=2) abline(v = statObs) text(10, 10, "observed corelation = ") } list( obs = statObs , sumFit = sum(statSim &gt; statObs)/numR) } </code></pre> <p>In fact it is hard for me to belive that correlation coefficient between two original matrices is high, and the one between the first original matrix and the second re-sampled one is maximal 0.2 after 10000 bootstrap repetitions. </p> <p>Any comments on the validity of the code?</p>
 

Querying!

 
Guidance
An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload