Exploring the limits of HRG for network modeling
16Nov16 |toy.py
reading cindy rules fails in g.sample
Walk throug the parsing it might do the thing
09Jan16 |New sampler (in david.py) that should use much much less memory.
To print the the Tree:
in tree_decomposition
, line 202 (new_visit
)
To print the Prod Rules:
in tree_decomposition
, line 299 (add_to_prod_rules
)
karate_chop.py
Working script used for developmenttw_karate_chop.py
working script used to generate visuals
HRG offers two methods of generating graphs:
- Exact Generation
- Stochastic Genration
- There is also Probabilistic HRG
Status | Description |
---|---|
Closed | Made a mod to weighted_choice function; the problem is n_rhs (line 432) some times the returned string is a tuple ... check the normal case. |
- To switch to the dev version of networkx, make PYTHONPATH
/home/username/local/python/lib/python2.7/site-packages/
- http://stackoverflow.com/questions/24364770/random-sampling-from-a-list-based-on-a-distribution
- On Assorativity
- Learning Bayesian Networks from Data | Nir Friedman Daphne Koller
Date | Notes |
---|---|
19Jun17 | this works: python pami.py --orig ../datasets/out.as20000102 |
29Jun17 | got phrg, cl, and kp working on small grpahs 29Jun17 | Next: work on larger graphs 02Jul17 | Begin Time experiments and large datasets 04Jul17 | Use the nu metrics to do stats 04Jul17 | working on net stats 04Jul17 | nu_metrix.py pami.py, need to reconsider the CDF 05Jul17 | Working on ECDF merging each result on x " 05Jul17 | Working on ECDF merging each result on x 05Jul17 | Workin on generating dat for large datasets 05Jul17 | ECDF http://ars.els-cdn.com/content/image/1-s2.0-S1877750315300259-gr7.jpg 17Jul17 | ToDo: run both Probs and Rods" 17Jul17 | ToDo: run both Probs and Rods
16Nov16 |toy.py
reading cindy rules fails in g.sample
Walk throug the parsing it might do the thing
09Jan16 |New sampler (in david.py) that should use much much less memory.
To print the the Tree:
in tree_decomposition
, line 202 (new_visit
)
To print the Prod Rules:
in tree_decomposition
, line 299 (add_to_prod_rules
)
karate_chop.py
Working script used for developmenttw_karate_chop.py
working script used to generate visuals
HRG offers two methods of generating graphs:
- Exact Generation
- Stochastic Genration
- There is also Probabilistic HRG
Status | Description |
---|---|
Closed | Made a mod to weighted_choice function; the problem is n_rhs (line 432) some times the returned string is a tuple ... check the normal case. |
- To switch to the dev version of networkx, make PYTHONPATH
/home/username/local/python/lib/python2.7/site-packages/
- http://stackoverflow.com/questions/24364770/random-sampling-from-a-list-based-on-a-distribution
- On Assorativity
- Learning Bayesian Networks from Data | Nir Friedman Daphne Koller
Date | Notes |
---|---|
19Jun17 | this works: python pami.py --orig ../datasets/out.as20000102 |
29Jun17 | got phrg, cl, and kp working on small grpahs
29Jun17 | Next: work on larger graphs
02Jul17 | Begin Time experiments and large datasets
04Jul17 | Use the nu metrics to do stats
04Jul17 | working on net stats
04Jul17 | nu_metrix.py pami.py, need to reconsider the CDF
05Jul17 | Working on ECDF merging each result on x "
05Jul17 | Working on ECDF merging each result on x
05Jul17 | Workin on generating dat for large datasets
05Jul17 | ECDF http://ars.els-cdn.com/content/image/1-s2.0-S1877750315300259-gr7.jpg
17Jul17 | ToDo: run both Probs and Rods"
17Jul17 | ToDo: run both Probs and Rods
25Jul17 | gcd taking for ever
31Aug17 | remote login \ Push URL: git@github.com:abitofalchemy/hrg_nets.git
31Aug17 | working dir: saguinag@dsg1:/data/saguinag/hrg_nets$
01Sep17 | working on exact hrg to save graphs
03Sep17 | set max loops taking a long time!
05Sep17 | got cl_kron_synth.py working to gen pickle files to be used with pgd
05Sep17 | todo: summarize the macro files for synth graphs
07Sep17 | 5546444797 Sep 5 10:18 Results/amazon0312_clgms.pickle, 644108189 Sep 6 09:29 Results/amazon0312_hstars.pickle
15Sep17 | todo: add a ret_val and return overall
15Sep17 | todo: in net_metrics
15Sep17 | fast exact phrg and net metrics can do faster clust
16Sep17 | got fast exact phrg working