[mlpack-git] [mlpack/mlpack] Modeling LSH For Performance Tuning (#749)

[Ryan Curtin]

> +  // the arithmetic and geometric mean of distances from each anchor to its
> +  // k-Nearest Neighbor.
> +  // The geometric mean of N numbers is the Nth root of the product of the
> +  // numbers. Through logarithmic properties though, this becomes computable
> +  // through exponentiating the mean of the logarithms of x:
> +  // exp(mean(log(x))) = geometricmean(x).
> +
> +  // Number of samples to create for modeling the Gamma Distributions
> +  size_t regressionExamples = 50; // TODO: parameter?
> +
> +  // Number of points to use as queries.
> +  size_t numAnchors = (size_t) std::round(0.1 * numSamples);
> +  arma::mat queryMat = sampleSet.cols(0, numAnchors - 1);
> +  // Evenly spaced sample sizes.
> +  arma::Col<size_t> referenceSizes = arma::conv_to< arma::Col<size_t> >::from(
> +    arma::linspace(numAnchors, numSamples - numAnchors - 1,

Can't you call linspace with `arma::Col<size_t>` as a template parameter so you can avoid the `conv_to`?  I think that is possible.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/749/files/cdcb575826bfb3bd0ef4cafacf465435b3d6d144#r74688857
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160813/3bcbca0a/attachment-0001.html>