Welcome to ArtDat
This folder contains code for generating artificial data with specified standard deviations. The nlfit direction file for\dfermi.dir is modified to include the beginning, end and number of data points. The data file name is used to place the resulting data.
The method for producing a Gaussian distribution is described in Random.doc.
../../../optimization/amoeba/ArtDat.doc ß earlier work on artificial data
../../SplineFitting/Feb24.doc ß refers to funwe.zip
Sum.doc – describes a simple method for going from the derivative to the integral. A fit to the sum is a reasonable way to produce a representation of the integral, but there is a “systematic” error introduced by the sum that can make the fit seem more accurate than in really is.
The code is for\gendat.for - for\GENDAT.WPJ. It can also be compiled by
Wfl386 gendat
The direction file is for\dfermi.dir.
aigau2.dir, OUTPUT FILE NAME
TEST.out, DATA SET NAME
1,1,
DIMENSION OF X IN F(X), power in ((f-fA)/err)**(2*npow)
C,0,1,0,
'R' read data point,err, 'C' err=SQRT(a+b|f|)+c|f|
5
NUMBER OF CONSTANTS IN POLY
400,4
MIN STEPS, Multi 0 NONE, 1_PPCC+MPH, 2 PPCC+MP, 3 P,PPCC,
.1,
INITIAL FR DECREASE DESIRED
1 VARY O FIX FITTED CONS
ERROR IN CONS
0 1000. ß
area of peak
1 -8 ß
essentially width of peak
1 360 ß location
of peak
1 .9 ß power of
Fermi function
1 0 ß
background
0,1000, beg, end off data range.
500,
Number of data points
Artificial data is stored in the file with the name given in the dir file.
The BLI is used to find an unevenly spaced data set with NP (from dir file) data points which best represent the underlying function. This data stored in the file requested for output with a UFN (underlying function) extension.
Eccentricities that can result from ‘over-fitting’ the data should be obvious in the underlying function.
Figure 1 Artificial data at intervals specified by bli.
Figure 2 Expanded view of region near peak.
zips
cpp\FUNWE.ZIP ß These are described in ../../../class07/FourierApr17.doc