## Matlab stats utilities

- Matlab stats utilities (see documentation and examples within the functions)

AffinityPropagationClusteringInterface.m | - | Cluster vectors using Frey's affinity-propagation algorithm. |

AssignStatsConstants.m | - | A script assigining several statistical constants used by other functions |

DisplayPCA.m | - | Display first principal components of data |

FindDistWithGivenKL.m | - | Compute a distribution with a given KL distance from P |

FindIndependentDistWithGivenKLFromUniform.m | - | Compute an indep. dist. with a given KL distance from uniform |

KL_distances_rand_points.m | - | Compute distances between many pairs of 'random' points (distributions). |

NextBestTree.m | - | Computes the 'next best' spanning tree |

NormalizeData.m | - | Normalization - Subtracts the mean of each row and divide by standard deviation (from Tal Shay) |

Phi.m | - | Standard Gaussian Phi |

PhiInv.m | - | Inverse standard Gaussian Phi |

PolishMean.m | - | Calculate mean after without lower/upper t |

add_noise_to_vec.m | - | Adds white gaussian noise to a vector |

bernoulli_sum_prob.m | - | Compute the probability distribution of sum of N independent bernoulli random |

cdf_hist.m | - | Compute cdf(t) for a histogram |

chinese_restaurant_process.m | - | Gerenate an instance of chinese restaurant process |

cond_mean.m | - | Compute conditional mean of a distribution given that value is > alpha |

convolution_prob.m | - | Return the convolution of two probabilities: p+q-2pq |

cumsum_hist.m | - | Perform cumulative summation for a histogram |

decide_success_or_failure_online.m | - | Internal function for deciding when to stop trying (if number of failures |

derich_correct.m | - | Add a relative derichlet correction |

displaysamplespca.m | - | Display a PCA of the samples. |

distributions_diff.m | - | Compute the difference between two distributions. |

distrnd.m | - | Simulate data from any distribution (like normrnd but for general dists.) |

diststat.m | - | Compute mean and var of a general distribution |

enrichment_fdr_plot.m | - | Plot the 'enrichment' FDR when comparing two sets, |

entropy.m | - | Compute entropy (in log base 2) |

ewen_sampling_formula.m | - | Generate a sample from Ewen's distribution |

gauss_smooth.m | - | Simple 1d Gaussian smoothing of a histogram. |

generate_points.m | - | Generate a set of points on the plane |

h.m | - | Binary entropy function |

hypergeometric_for_many_sets.m | - | Give hypergeometric-like p-value for pairwise intersections of many sets |

hypergeometric_for_three_sets.m | - | Give hypergeometric p-value for 3-way intersection, given pairwise intersections |

indian_buffet_process.m | - | Gerenate an instance of chinese restaurant process |

integral_hist.m | - | Perform integral for a histogram (we use simple rectangle integration) |

integral_hist2d.m | - | Perform two-dimensional integral for a histogram (we use simple rectangle integration) |

isbimodal_hist.m | - | Test if a distribution is bimodal |

kurtosis_hist.m | - | Compute kurtosis for a histogram |

log_binom.m | - | Compute log of binomial coefficient N over M. |

log_factorial_vec.m | - | Compute log-factorial for a vector. |

maxnormcdf.m | - | Cumulative distribution function of maximum of N gaussians |

maxnormpdf.m | - | Density of maximum of N standard Gaussians |

maxnormstat.m | - | Compute mean and st.d. of maximum of N standard Gaussians |

mean_hist.m | - | Compute mean for a histogram |

mean_not_nan.m | - | unction vec = mean_not_nan(mat) |

mean_single.m | - | A function for computing the mean of a large single array that |

med_hist.m | - | Compute median for a histogram |

median_hist.m | - | Compute median for a histogram |

median_mad.m | - | unction [median_vec, mad_vec] = median_mad(mat) |

moment_hist.m | - | Compute central moment of any order for a histogram |

my_cov.m | - | Compute covariance for two matrices |

my_least_squares.m | - | Find the best a and b such that the sum of squares of y - (ax+b) is minimized |

my_quantile.m | - | Like Matlab's quantile but enables to get quantiles outside [0,1]. |

my_smooth.m | - | Performs moving sum (default)/average |

ncx2power_to_ncp.m | - | Compute the non-centrality parameter needed to achieve a desired power |

normalize_hist.m | - | Normalize a histogram to have sum one |

normalize_hist2d.m | - | Normalize a two-dimensional histogram to have sum one |

plane_check.m | - | Check if certain points are on the same plain |

poisson_point_process_rnd.m | - | Generate data from a Poisson point process in the cube [0,1]^n |

posdefrnd - Copy.m | - | Sample a semi positive-definite matrix |

posdefrnd.m | - | Sample a semi positive-definite matrix |

powernormstat.m | - | compmute moments of |z|^p when p is standard Gaussian |

pvalPearson.m | - | VALPEARSON Tail probability for Pearson's linear correlation. |

quantile_hist.m | - | Compute quantile for a histogram |

quantile_normalize.m | - | Perform quantile normalization on a matrix. |

rand_nchoosek.m | - | Pick at random k indices out of n |

rand_normalized.m | - | Randomize an mXn matrix such that each column sums to 1 |

rand_seed.m | - | A script for seeding Matlab's rand |

randcircle.m | - | Draw random points in a circle |

random_Tree_to_point.m | - | Get the closest point on a random tree (in KL) to a point (distribution). |

randsphere.m | - | Generate random points inside a 3-d sphere |

ranksum1side.m | - | Wilcoxon rank sum test but with one-sided p-value |

rationormpdf.m | - | Ratio of two Gaussians density distriburion |

rationormstat.m | - | First two moments of ratio of gaussians |

relative_entropy.m | - | Compute the relative entropy between distributions P and Q |

sample_gaussian_integral.m | - | A Gaussin integral |

set_windows_pval.m | - | Assume n variables are uniformly distributed on [1, N] |

set_windows_pval_old.m | - | Assume n variables are uniformly distributed on [1, N] |

simple_binom_windows_pval.m | - | Assume n variables are uniformly distributed on [1, N] |

simulate_by_normalization.m | - | Simulate uniform r.v.s. normalized by their sum |

simulate_correlated_bernoulli.m | - | Simulates a set of correlated bernoulli random variables |

simulate_correlated_gaussians.m | - | Simulates a set of correlated gaussian random |

simulate_quasi_regression.m | - | Simulate linear regression model and quasi regression |

simulate_rand_points.m | - | A script for simulating throwing random points in the unit interval |

skewness_hist.m | - | Compute skewness for a histogram |

std_hist.m | - | Compute std for a histogram |

std_to_q_binary.m | - | Compute Bernoulli probability for a given standard deviation |

sum_hist.m | - | Add values in two histograms |

table_to_marginal_probs.m | - | Compute product of marginals for a probability table |

test_KL_distances_rand_points.m | - | Script for runnning the test_KL_distances_rand_points function |

test_MI_computation.m | - | Test inference and mutual information computation for trees. |

test_chisquare.m | - | Test for conditional indepndence via chi-square test (from Kevin Murphy's BNT) |

test_correlated_gaussians.m | - | Generate correlated gaussians. Then perform de-convulution |

test_distributions_diff.m | - | A script for testing the distributions_diff function |

test_goodness_of_fit.m | - | function test_goodness_of_fit() |

test_hypergeometric_for_many_sets.m | - | Test the function hypergeometric_for_many_sets |

test_hypergeometric_for_three_sets.m | - | Give p-value for the hypergeometric score of intersection of three sets, |

test_hypergeometric_scores.m | - | Test all hypergeometric functions |

test_probit.m | - | Run a test of matlab probit regression. Determine coefficients, simulate |

test_regression.m | - | unction test_regression |

test_windows_pval.m | - | Test the p-values for significance, using both windows methods, |

uniform_log_sum_reciprocal_mean.m | - | Compute the mean of -1/(sum_{i} log(p_i)) |

uniform_spanning_tree.m | - | Generate a uniform spanning tree on n vertices |

uniform_sum_reciprocal_mean.m | - | Compute the mean of 1/(sum_{i} p_i) |

var_hist.m | - | Compute variance for a histogram |

weighted_hist.m | - | Builds a histogram by binning the elements of val into containers, one |

weighted_mean.m | - | Compute weighted average |

weighted_rand.m | - | Produce random numbers drawn according to some weights. |

stats.tgz |
- | Download the whole package. (107 .m files) |

fdr |
- | Go to the fdr package. |

hmp |
- | Go to the hmp package. |

mog |
- | Go to the mog package. |