emblem
Estimate from sums
Attempts to estimate distribution parameters from sums of samples.
2024.Jul.03 18:29:26
Sample sizes
Equalize them ? Make sample sizes equal (if sum is divisible). •
μ, σ (Gaussian) Parameters (for RNG). •
α % Level of significance. •
.seed × Seed for random numbers. •

Attempts to estimate the parameters of the underlying Gaussian distribution from the given samples. The only information used is the sum of each sample, although the samples' data are simulated (to spare data input). Indeed, in the typical industrial applications, usually only the sums are available, not the individual values.

An option to equalize the sample sizes is provided, for simplicity. (The sum of the sample sizes must be adequately divisible.) An attempt is made to estimate σ, needing further (thorough) research.

Other suggested data: do not equalize.

References: Plate: EstimateFromSums

• #Dartmouth College, 2014, Chapter 2.pdf, "Continuous probability densities"

• #Weisstein, Eric W., "Disk Point Picking", MathWorld—A Wolfram Web Resource

• 1871-01-07: Borel, Félix Édouard Justin Émile (1956-02-03).

 
 
Valid HTML 4.01! IST http://web.ist.utl.pt/~mcasquilho/compute/qc/Fx-old_estimFromSums.php
Created: 2014-01-07 — Last modified: 2014-01-12