Welcome to my R package for interferogram analysis and manipulation of Zernike polynomials. If you are new to R this README gives installation and basic usage instructions for Windows users. There are a number of required and optional packages to be installed, so please follow these instructions carefully. You need:
-
A recent version of R itself (v4.1.0 or later is required). The current version at this writing is 4.4.2. If you're making a fresh installation just go with the current release. The windows binary installer can be downloaded from https://cran.r-project.org. Installation works like any Windows program -- I like to have the installer create desktop shortcuts and edit its properties to have it start up in the directory where I've stored data. The only customization I make at install time is to select "SDI" windowing mode. This will display graphs in separate windows on the desktop.
-
Download the Windows binary of this package from the releases section. Do not unzip it!
-
The following steps are done within R itself. When you start R a console window opens with a small number of menu items and a text entry area. Commands are typed at the ">" prompt. You should have an active internet connection for the following steps:
-
Install the following packages. This can be done with the menu item
Packages/Install package(s)...The first time you select this in a session it will prompt you for a mirrored download site. Next it will display a selection box with all available packages. The ones you need are:-
Rcpp. This package also has a large number of dependencies that should be automatically downloaded and installed.
-
RcppParallel.
-
RcppArmadillo.
-
BH.
-
-
Optional but highly recommended packages
-
jpeg, tiff, png. As of version 3.8.0 these are required to read interferogram image files of the respective types (previously "home grown" code was used to read jpeg or tiff files). Most raw format image files can be read with no additional package requirements.
-
rgl. This package provides interactive 3D graphics.
-
robustbase. The function
nlsrobis used as the new default parameter estimation routine by the automated edgefinding functioncircle.pars. -
tinytable. Used to produce html formatted reports for the fringe analysis routines.
-
fftwtools. If installed the FFT based fringe analysis routines
vortexfitandfftfitwill usefftwfor FFTs instead of the base R function. This may slightly improve the execution time of these functions.
-
-
You need not install these initially but they may be useful for certain tasks.
-
clue. The function
solve_LSAPis used by the branch cut algorithm for phase unwrapping implemented in the functionbrcutpuw. This may outperform the default phase unwrapping routine in some situations. -
data.table, dplyr, pixmap, mvtnorm.
-
my package lppuw, available on github at github.com/mlpeck/lppuw, which contains some advanced methods of phase unwrapping. This also requires a package named rcbc. A Windows binary is provided in the releases section of
lppuwfor anyone who needs it.
-
-
Now install package "zernike". This can be done with the menu item
Packages/Install from local zip file...At the prompt just navigate to wherever you saved the zip file and select it. If it installs successfully a brief message will be sent to the console. You may get a warning message if you're running a different version of R from the one the package was built in. You can probably ignore this -- if the demos run you're in good shape. -
Three demo programs are included with the program. To run them enter the following commands:
demo(psiest, package="zernike", ask=FALSE)
demo(winfit, package="zernike", ask=FALSE)
demo(pcaest, package="zernike", ask=FALSE)
These demos illustrate basic PSI analysis, a sliding window analysis of a multiple cycle PSI sequence, and some alternative algorithms for generalized phase shifting interferometry.
-
The main PSI analysis routine is
psifit(). Typeexample(psifit, package="zernike", ask=FALSE, echo=FALSE)to see an example of its use. This uses the same data set as the demo programs. -
I now have substantial examples for several high and not so high level functions in the package. Besides
psifit()these includevortexfit(),fftfit(),wf_net(),zpm_cart(),zapm(),zapm_iso(),zapm_128(), andzapm_iso_128(). These are run in the same way as described forpsifit()above. -
To access R's help in html format type
help.start(). This will open up a window in your default browser with a help page containing links to the help files for all installed packages on your system as well as manuals, FAQs, and help files for R itself. -
You can exit R by typing q() or just closing the console window. You will be prompted whether to save the workspace. If you enter y any data in your workspace will be saved in binary format to a file named .RData, and commands entered in the current session will be saved to an ordinary text file named .Rhistory.
This package runs very well in Linux, in fact it is developed on a PC running Fedora Linux. Building from source is straightforward provided you have appropriate development tools and libraries installed. It also runs in WSL on Windows 11, or with a little more effort even Windows 10. I have no access to Mac's and therefore can't help with installation on MacOS.
There is a vast quantity of documentation and literature about the R system. At a minimum you should read the R for Windows FAQ and the introduction to R included as a PDF file in the software installation.
PSI data are courtesy of Vladimir Galogaza. Steve Koehler provided valuable programming advice.