Index of /grad-school/bridge/walk_count_v2
Name Last modified Size Description
![[PARENTDIR]](/spicons/back.gif)
Parent Directory -
![[ ]](/spicons/unknown.gif)
Makefile 2010-02-11 12:11 739
![[ ]](/spicons/layout.gif)
atheta.pdf 2010-03-27 12:56 18K
count_walks 2010-04-19 00:58 14K
![[TXT]](/spicons/text.gif)
count_walks.c 2010-04-06 13:16 2.0K
count_walks.mk 2010-02-11 12:11 1.0K
count_walks.mki 2010-02-06 00:31 146
curvefit.aux 2010-03-26 12:08 452
curvefit.log 2010-03-26 12:08 9.3K
curvefit.out 2010-03-26 12:08 0
curvefit.tex 2010-03-26 12:08 2.6K
![[DIR]](/spicons/folder.gif)
data.bak/ 2010-04-16 12:16 -
data/ 2010-04-19 00:20 -
data1224/ 2010-04-19 00:45 -
databig/ 2010-04-21 18:41 -
datasmall/ 2010-04-16 12:38 -
fmtplot 2010-03-26 13:19 155
fmtplotnolog 2010-03-26 15:15 149
![[ ]](/spicons/p.gif)
hv_bonds.py 2010-04-19 01:09 8.4K
ice/ 2010-04-16 14:12 -
lattice_effects.tex 2010-04-07 12:42 8.7K
logcorrectfit.py 2010-04-06 21:38 8.6K
logcorrectfit2.py 2010-04-06 21:21 4.0K
logfit.py 2010-03-26 12:11 1.2K
make_raw_counts.py 2010-04-19 00:45 1.0K
objs/ 2020-07-26 21:46 -
old/ 2010-04-06 21:18 -
plot2 2010-04-16 15:25 88
powerlawminusfit.py 2010-04-19 00:32 2.0K
regression/ 2010-03-10 22:56 -
scale_raw_counts.py 2010-04-18 23:25 3.4K
sle_partition_func.tex 2010-04-07 12:44 8.1K
syms.tex 2010-03-26 12:04 7.8K
tags 2010-02-21 12:21 2.7K
theta.txt 2010-04-19 00:04 376
timing-info/ 2010-03-10 22:56 -
w_N_with_theta_serie..> 2010-04-19 01:07 10K
w_theta_with_N_serie..> 2010-04-19 01:07 8.0K
walk_count_lib.c 2010-04-19 00:58 20K
walk_count_lib.c.bak 2010-04-19 00:39 20K
walk_count_lib.h 2010-04-06 13:17 5.5K
================================================================
John Kerl
kerl.john.r@gmail.com
2010-02-03
================================================================
================================================================
DESCRIPTION
This is code for exhaustive enumeration of constrained walks on the 2D integer
lattice for small N (< 25 or so). This is not the same as MCMC methods which
work for much larger N, but which only sample instead of enumerating
exhaustively. This code uses a recursive backtracking algorithm to find all
walks of length <= N. Constraints are specified on the command line (see the
next section): self-avoidance constraints include none (e.g. simple random
walk), previous walk only (immediate-backstep avoidance), or full
self-avoidance; location constraints include full-plane, upper half-plane,
upper right quadrant, or Shane Passon's rational-slope constraint. Please see
walk_count_lib.h for more information.
================================================================
USAGE INSTRUCTIONS
Type "make". Then:
count_walks {constraints name} {N} [options]
or:
count_walks {constraints name} {Nmin-Nmax} [options]
Options: "kerl" or "shane" for output styles; "stdout" or "stderr" to print
individual walks.
E.g.
count_walks angle=3/7 1-12
count_walks UPSAW 1-12
count_walks UPSAW 1-5 stdout
Or: Just type "shane1.sh".
================================================================
TEST CASES
Here are test cases I have run. There are three self-avoidance constraints
and three location-conditioning constraints (besides Passon's slope
constraints). Some combinations of those have well-known names:
* No self-avoidance in the full plane: this is a simple random walk
("SRW"). There are 4^N such walks, since at each step one can choose any
of the four directions.
* No self-avoidance in the upper half-plane: this is the half-plane
excursion (HPE).
* No self-avoidance in the upper right quadrant: this shows up in Sloane's
integer sequences.
* Avoiding only immediate backsteps in the full plane: there are 3^N such
walks, since at each step one can choose from any of the three
non-back-step directions.
* Avoiding only immediate backsteps in the upper half-plane or upper right
quadrant: I haven't found references to these anywhere.
* Self-avoiding walk in the full plane: "SAW". Counts are available on the
web, including via Sloane.
* Self-avoiding walk in the upper half-plane: "UPSAW". Counts are
available on the web, including via Sloane.
* Self-avoiding walk in the upper right quadrant: I don't know of a short
name for these. Counts are available via Sloane.
Neil Sloane's sequences may be found at
http://www.research.att.com/~njas/sequences,
and/or Google for OEIS (On-line Encyclopedia of Integer Sequences).
The following table summarizes my results, using this program, for these
known-value test cases. Counts are shown for N=0,1,2,3,... . What I am
showing here is the correctness of my code: for each case where there were
results to compare to, I indeed obtained the correct results.
Slf-avd Loc Desc. Verif. Count
------- --- ---- ------ -----
None FP SRW Yes 4^N 1,4,16,64,256,1024,4096,16384,65536
None UHP HPE Yes Sloane A001700 1,1,3,10,35,126,462,1716,6435,24310
None URQ ??? Yes Sloane A005566 1,2,6,18,60,200,700,2450,8820,31752
Prev. FP ??? Yes 4*3^(N-1) 1,4,12,36,108,324,972,2916,8748,26244
Prev. UHP ??? N/A ??? 1,1,3,7,19,53,147,411,1163,3313,9475
Prev. URQ ??? N/A ??? 1,2,4,10,26,66,174,462,1256,3410,9410
Full FP SAW Yes Sloane A001411 1,4,12,36,100,284,780,2172,5916
Full UHP UPSAW Yes Sloane A116903 1,1,3,7,19,49,131,339,899,2345,6199
Full URQ ??? Yes Sloane A038373 1,2,4,10,24060,146,366,912,2302,5800