Ceci est une ancienne révision du document !
I have a dedicated R program for processing data. So, I have downloaded R, and the associated R-studio for my work laptop. The R-studio is a GUI interface for R and is divided into 4 panes. The upper left is the R program that you can import from Leafpad. The lower left is the actual real-time command-line processing. The upper right is publishing the rights for the statistical outputs and tables for journal abstracts. The lower right is the produced tables.
R has the capability of creating a Shiny app which is an online capability for the R programs that you use. It is now possible to be connected to your R programs using a web browser. The Shiny app replaces a previous Perl Batch program. The batch program required 2 hrs, the Shiny app does the same amount of work in 20 minutes.
The Shiny app then generates a pdf of the waveform plots (as seen in the lower right) and the associated critical points. The Shiny app works well, but there are a few “bugs” in the R-code written by my biostatistician so the app is not perfect; however it saves time and removes technician bias in data processing. Yet it does a great job of generating the data spreadsheets for the pressure and motion data sets needed. An example spreadsheet below.
Onto gnuplot now.
I read the starting pages on a rather dry subject: statistics. There are 2 software developers’ forewords, an ‘about this book’, 15 chapters, and an appendix. The starting paragraphs have important background and historical information on Gnuplot. Yet it does not cover the real story. I jumped into Chapter 1, and it is quite light with a 15 page span.
Luckily, Gnuplot is part of the RPM and Debian repositories. I changed to root and installed gnuplot via terminal – which was an incredibly smooth process. I searched the menus and did not see an app icon. I rebooted, right-clicked, and launched terminal with gnuplot. My book is written about Gnuplot 4.0, not 5.0. I am banking on the idea that gnuplot is relatively static and that the commands are reliable.
Chapter 1 merely reviews the scope and capabilities of Gnuplot. It a simple 15 pages that goes into a brief description of commands. The authors use an example of planning a morning marathon and staffing issues. Essentially, it was attempting to use bimodal statistics to highlight a need that there were two surges of marathon runners: professionals and amateurs. The staff would needed to be present at early start, say 10am, then 11am and 1pm. The professionals would end at 11 am while the amateurs will be 1 pm. Staffing would be heavy at those times. This chapter ends by stating that each chapter will treat the reader as a new user. I believe the final message is that gnuplot will “illuminate” the truth found in statistical data.