MAKCSA: help and advice

MAKCSA: help and advice


Consult the reference manual for further details and worked examples.
W.G.Bardsley, University of Manchester, U.K.
Program operation

If an experiment generates a set of n observations that have been ordered such that

x1 =<   x2 =<   x3 ... =<   xn

then a histogram consists of the frequencies with which the observations accumulate within m cells defined by m sets of lower limits Li and upper limits Ui where

L1 =<   x1, ..., xn =<   Um

L1 <   U1 = L2 <   U2 = L3 ... <   Um.

This program simulates data for such histograms in order to study the fitting of probability models to histogram data types.

In particular it simulates flow cytometry data in order to explore use of the program CSAFIT, which uses spline smoothing then nonlinear regression to fit flow cytometry profiles according to the linear shift and stretch model. This is how to use the program.

  1. Input the total number of cells to be counted in the data
  2. Input the total number of channel settings for histograms
  3. Input the first and last histogram channel fluorescences
  4. Input the mean (mu) and standard deviation (sigma) to use when simulating data sets from the normal distribution
  5. Input the % stretch and % translation that you require
  6. Input the % relative error that you want to add to data
  7. Display current histograms of X and Y with added error
  8. When you are happy with a set of data, write it to a file
  9. Take the output file and analyse it using program CSAFIT

The model used in this program is a normal distribution with parameters mu and sigma. The error added is a % constant relative error, that is a a constant coefficient of variation. Default settings are provided (to help you get started) and you should try beginning with these set values. For details see

Bardsley et al J. Immunol. Meth. 153, 235-247 (1992) and

Bardsley and Kyprianou J. Math. Biol.(1996) 34:271-296.