Basics on Statistical Analysis
Some notes on basic conceps and it tools in R and some examples on ROOTR
Distribution/pdf | Example use in HEP
|
Binomial | Branching ratio
|
Multinomial | Histogram with fixed N
|
Poisson | Number of events found
|
Uniform | Monte Carlo method
|
Exponential | Decay time
|
Gaussian | Measurement error
|
Chi-square | Goodness-of-fit
|
Cauchy | Mass of resonance
|
Landau | Ionization energy loss |
http://www.pp.rhul.ac.uk/~cowan/stat_cern.html
erf(Error Function) and erf.inv (Inverse Error Function)
http://mathworld.wolfram.com/InverseErf.html (For Gaussian Distribution)
erf <- function(x) 2 * pnorm(x * sqrt(2)) - 1
erf.inv <- function(x) qnorm((x + 1)/2)/sqrt(2)
Example
erf.inv(1)
erf.inv(0)
erf.inv(-1)
erf.inv(erf(.25))
erf(erf.inv(.25))
erf.inv(.5)
source
link
KDE Kernel Density Estimation 1D
The way to get PDF ;D
Just use de function density(x)
PDF (Probability Density Function)
http://reference.wolfram.com/mathematica/ref/SmoothKernelDistribution.html
http://reference.wolfram.com/mathematica/ref/PDF.html
x <- log(rgamma(150,5))
PDF <- approxfun(density(x))
plot(density(x))
xnew <- c(0.45,1.84,2.3)
points(xnew,df(xnew),col=2)
source
link
Multivariate Analysis Approach
KDE Kernel Density Estimation
Exist a package called ks that lets you to calculate a KDE and KDA
letting you to calculate an optimal h(bandwidth or smoothing) with full or diagonalized matrix.
paper at
http://www.jstatsoft.org/v21/i07/paper
Autor: Omar Andres Zapata Mesa