Frida (“Fast reliable interactive data analysis”) is a versatile data analysis program with special routines for inelastic neutron scattering. It is actively maintained by Joachim Wuttke.
11aug17: Release 2.4.0g. Corrected ticks and tacks, improved plot symbols, improved i96 loader.
21apr17: Release 2.4.0b. Bug fix for fit weights. New functions sqrt0, ln0, lg0 return 0 where sqrt, ln, lg now return NaN.
24mar17: Release 2.4.0a. PostScript definitions simplified and reorganized in file wups17a.ps. Support for 2D plots (only through PostScript, not through gnuplot). Bug fix reenables index computations like “y[,,ni-1-i]”.
17mar17: Release 2.3.6c. Binning reverted to old default; computation of sampling variance can now be requested through option 's'. Several minor bug fixes, and improvements of functionality. New dependency: boost/filesystem.
1feb17: Release 2.3.5a. Improved CMake and CTest machinery. Other recent changes: commands mpgi (interpolation) and mpg (histogram redistribution); command fthb to read blocks with no header; command cfs restored; loaders for i96 and a01 restored.
18may16: Lecture notes on post-processing Frida graphics available at fridagraphics.pdf.
20oct15: Release 2.3.2a comes with a new data format, with file name extension .yda. Users won't notice much of a difference; old .y08 will of course still be read. Furthermore, the curve evaluation syntax has changed from
f[K,J](T). New online doc commands hc and hv, some improvements in internal workings.
22may2015: Frida has a new list format. The list format is ubiquitous in Frida, since it is used to select online files or spectra within a file. From now on, lists must be given in the form
from:to:step instead of
from-to.step. The full specification is
[from]:[to][:step], i.e. everything besides the first colon can be omitted and will be filled with default values. So just say
: for a full list (old format:
5: for a list extending from 5 to the end (old format
5-*). This change will require a major change of habits, and therefore cause some inconvenience. However, it will open up new paths for the future evolution of Frida that were blocked by limitations and side effects of the old syntax.
02feb2015: Since release 2.2.3a, Frida supports global fit parameters. Also some of the innermost mechanisms of Frida have been overhauled to make the arithmetics type aware. The most obvious immediate advantage is that index computations are now carried out entirely in the integer domain. Just use the integer division operators
18dec2014: With release 2.2.1b, the command
oio is finally gone. Remember that the second 'o' stood for 'old'; this command was since long meant to be phased out. To replace
oio, there is now a suite of functions that operate on a spectrum and return a scalar, like
valmin etc. For instance, to compute the sum of a histogram, use
oi sum. To compute an integral, use
oi integral. For help, use
See the changelog for details.
Please send bug reports to Joachim. The git development snapshot contains a bug list (~/TODO).
The first JCNS School on Data Analysis for Quasielastic Neutron Scattering with Frida took place on 28-29 April 2010. The next school will be organized as soon as a sufficient number of potential participants have indicated their interest.