Sunday, February 28, 2016

BML Simulator Batch Mode

This note announces release of a "batch mode" upgrade to the Binary Mechanics Lab Simulator (BMLS) v1.39 which may be downloaded by clicking the link. In addition to the hotspot 1.39 simulator, the download contains five *.bat files in its root directory (mine is c:\physics\hotspot) and a \bat subdirectory containing five examples of input parameter files in Microsoft text format (lines delineated with carriage return {13} and line feed {10}, 0D 0A sequences when viewed in hex format).

Fig. 1: Input Parameter File Format


Upon startup, the BMLS looks for a command-line parameter specifying the pathname of an input parameter file, formated as shown in Fig. 1. If this is not null, the BMLS looks for the file. In the download configuration, these files are located in a \bat sub-directory. If the file is found, it is used for input parameters and runs to completion without further user intervention.

Five *.bat files are provided in the root directory: ex01.bat, ex02.bat, gravity.bat, mag_up.bat and mag_down.bat. Each file contains a single command-line to run the BMLS using a particular parameter input file. For example, ex01.bat contains one line:

c:\physics\hotspot\hotspot.exe bat\ex01.txt

The sample *.bat files should work without any editing if the download is unzipped into a c:\physics\hotspot directory. Any number of such command-lines may be included in a single *.bat file to automate runs of multiple experiments including repeating a run with different initial random states to develop a number of trials for statistical analysis. In the following listing of ex01.txt,
56L220suvf01Ex1
0
56
1
11
1,0,0,0
0.22
255
255
0
100
N
1
suvf
5000
a file-naming convention is illustrated. The filename in the first line includes: Dim=56, a capital letter code (L, laser; B, Box mode; R, Rnd mode; V, Vacuum mode, etc), 3 digits of initial bit density, bit operations order (e.g., suvf), trial number (01), experiment designation if applicable (Ex1). The second line is 0, indicating random seeding of the initial state based on the CPU timer value. A specific non-zero integer here would (should!) repeat the same "random seeding" each time. Notice that if the trial number 01 were incremented (02, 03, etc), with the second line 0, a number of runs could be automated with the only change being output file name and the random seeding of the initial state.

ex01.txt and ex02.txt (featured in [1]) contain just one iteration of the random seeding of the initial state for the entire simulated volume. The other three examples include multiple iterations where different sub-volumes were initialized or different bit masks used to control which spot types or bit types would be initialized with the specified probabilities (line 7). There is no limit to the number of iterations of lines 7 to 12, and hence, no limit to the complexity of the initial states created.

The selection of the spot type, bit type and spot location parameters allow for initializing a single bit in the 1-state. That is, the resolution is the maximum possible. For example, these input parameters were used to initialize a single 1-state lite bit in the proton bit cycle [2] in a volume otherwise "filled" with absolute vacuum (the 011glu1.mat file in the \ini subdirectory), showing that a proton has infinite life-time in absolute vacuum [3] without any external electromagnetic fields.

To try out batch mode, just double-click on one of the *.bat file examples.

Discussion
BMLS [4] batch mode might be welcome to all users tired of manually typing input parameters. The next logical development would be a simple Windows GUI program where one could select input parameters mostly with point-and-click, which would write the input parameter files and *.bat files or simply launch the BMLS with the appropriate command-line.

Another advantage of batch mode might be that users have a record of exactly what was done in the form of small text files that can be archived or shared with other investigators, essentially specifying the methods used. More than once, the author has observed very interesting results but found that the effects occur only when specific input parameter configurations are used, and then ..., the inputs are not recalled exactly to replicate the effect.

What might be the top priority for users? Show that something is wrong with binary mechanics fundamentals. "3, 2, 1, GO".

References
[1] Keene, J. J. "Electron gas standing waves" J. Bin. Mech. February, 2016.
[2] Keene, J. J. "Proton and electron bit cycles" J. Bin. Mech. April, 2015.
[3] Keene, J. J. "Vacuum thresholds" J. Bin. Mech. March, 2011.
[4] Keene, J. J. "Binary Mechanics Lab Simulator update" J. Bin. Mech. December, 2015.
© 2016 James J Keene