Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Gawron default values inconsistency between duarouter/documentation and duaIterate.py #14772

Closed
jonasjostmann opened this issue Apr 25, 2024 · 4 comments

Comments

@jonasjostmann
Copy link

The documentation of the duarouter (main/docs/web/docs/duarouter.md) and the script itself (/src/duarouter/RODUAFrame.cpp) use these default values for the Gawron parameters:

Option Description
--gawron.beta {{DT_FLOAT}} Use FLOAT as Gawron's beta; default: 0.3
--gawron.a {{DT_FLOAT}} Use FLOAT as Gawron's a; default: 0.05

The main/tools/assign/duaIterate.py, however, uses 0.5 for the alpha and 0.9 for the beta:

argParser.add_argument("-A", "--gA", type=float, default=.5, help="Sets Gawron's Alpha")
argParser.add_argument("-B", "--gBeta", type=float, default=.9, help="Sets Gawron's Beta")

@namdre
Copy link
Contributor

namdre commented Apr 29, 2024

the values in duaIterate.py were never changed since the initial commit:
22ee2f8#diff-ef9ec561b42078d287fcc807b5d2ce5f00e5239bce86c248ef7e3a456b82f744R91-R94

the values in the C++ code were originally the same and then changed 21 years by a seemingly unrelated commit: 54bf796#diff-eca2ce9e02e8603941ac99a924eac40d71d7aef95b582316c6d9a74329dade9b

I suspect most people use the functionality through duaIterate.py so 0.5 and 0.9 are the more relevant numbers.

@namdre
Copy link
Contributor

namdre commented Apr 29, 2024

Daniel Krajzewicz recalls the reason for the change: Dr. Nils Eissfeldt told him over the phone that those values were better 😄
Daniel is happily using the duaIterate.py defaults along with everyone else, however.

@yunpangfloetteroed
Copy link
Contributor

yunpangfloetteroed commented Apr 29, 2024

The beta value, used in Gawron's dissertation, is 0.05 (p. 47), which prevents drivers from remembering forgetting the travel times of routes they have not used for a long time.

The parameter a affects the convergence performance, where a = 0.5, 1 and 2 were used in Gawron's dissertation (p.48). The larger the a, the faster the convergence. Formula (4.17) helps to check if a is desirable.

namdre added a commit that referenced this issue Apr 29, 2024
@namdre namdre closed this as completed in d4da879 Apr 29, 2024
namdre added a commit that referenced this issue Apr 29, 2024
@namdre
Copy link
Contributor

namdre commented Apr 29, 2024

the somewhat surprising test changes led to the discovery of #14790 and #14791

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants