POC-English-NoAmb tests `2018-10-03`¶

Connector identificators changed (2018-09-26): C01 ⇒ B/AB/AAB...
.dict file format updated to Link Grammar 5.5.1 -- Local ULL-DEV environment.

Each line is calculated 1x, parsing metrics tested 1x for each calculation.
The calculation table is shared as 'short_table.txt' in data folder
http://langlearn.singularitynet.io/data/clustering_2018/POC-English-NoAmb-2018-10-03/
This notebook is shared as static html via
http://langlearn.singularitynet.io/data/clustering_2018/html/POC-English-NoAmb-2018-10-03.html

Basic settings¶

import os, sys, time
from IPython.display import display
module_path = os.path.abspath(os.path.join('..'))
if module_path not in sys.path: sys.path.append(module_path)
from src.grammar_learner.utl import UTC
from src.grammar_learner.read_files import check_dir
from src.grammar_learner.widgets import html_table
from src.grammar_learner.pqa_table import table_row  # updated, incl. silhouette index
from src.grammar_learner.write_files import list2file
tmpath = module_path + '/tmp/'
check_dir(tmpath, True, 'none')
table = []
long_table = []
header = ['Line','Corpus','Parsing','LW','RW','Gen.','Space','Rules','Silhouette','PA','PQ']
start = time.time()
print(UTC(), ':: module_path =', module_path)

2018-10-03 17:11:06 UTC :: module_path = /home/obaskov/py/language-learning

Grammar Learner corpus-specific parameters¶

corpus = 'POC-English-NoAmb'
out_dir = module_path + '/output/'+ corpus + '-' + str(UTC())[:10]
runs = (1,1)    # (attempts to learn grammar per line, grammar tests per attempt)
if runs != (1,1): out_dir += '-multi'
kwargs = {
    'left_wall'     :   ''          ,
    'period'        :   False       ,
    'word_space'    :   'vectors'   ,
    'cluster_range' :   (50,2,9)    ,   # (2,30,1,5) - 5 is not enough
    'clustering'    :   ('kmeans', 'kmeans++', 18),
    'tmpath'        :   tmpath      , 
    'verbose'       :   'max'       ,
    'template_path' :   'poc-turtle',
    'linkage_limit' :   1000         ,
    'categories_generalization': 'off' }
lines = [
    [14, 'POC-English-NoAmb', 'MST-fixed-manually'              ,'LW','.', 'none' ],
    [15, 'POC-English-NoAmb', 'MST-fixed-manually'              , 0  , 0 , 'none' ],
    [16, 'POC-English-NoAmb', 'LG-English'                      ,'LW','.', 'none' ],
    [17, 'POC-English-NoAmb', 'LG-English'                      , 0  , 0 , 'none' ],
    [18, 'POC-English-NoAmb', 'R=6-Weight=6:R-mst-weight=+1:R'  ,'LW','.', 'none' ],
    [19, 'POC-English-NoAmb', 'R=6-Weight=6:R-mst-weight=+1:R'  , 0  , 0 , 'none' ],
    [20, 'POC-English-NoAmb', 'R=6-Weight=1-mst-weight=+1:R'    ,'LW','.', 'none' ], 
    [21, 'POC-English-NoAmb', 'R=6-Weight=1-mst-weight=+1:R'    , 0  , 0 , 'none' ], 
    [22, 'POC-English-NoAmb', 'LG-ANY-mst-weight=+1:R'          ,'LW','.', 'none' ],
    [23, 'POC-English-NoAmb', 'LG-ANY-mst-weight=+1:R'          , 0  , 0 , 'none' ]]
# cp,rp :: (test) corpus_path and reference_path
cp = module_path + '/data/POC-English-NoAmb/poc_english_noamb_corpus.txt'
rp = module_path + '/data/POC-English-NoAmb/poc-english_noAmb-parses-gold.txt'
#cp = rp  # July 2018 Grammar Tester option

ULL Project Plan ⇒ Parses Backup ⇒ lines 14-23¶

Connectors-DRK-Connectors¶

%%capture
kwargs['context'] = 1
kwargs['word_space'] = 'vectors'
kwargs['clustering'] = ('kmeans','kmeans++',18)
#kwargs['cluster_range'] = (50,2,9)  # thorough clustering search for max silhouette index
kwargs['grammar_rules'] = 1
average21, long21 = table_row(lines, out_dir, cp, rp, runs, **kwargs)
table.extend(average21)
long_table.extend(long21)

display(html_table([header]+average21))

Connectors-DRK-Disjuncts¶

%%capture
kwargs['grammar_rules'] = 2
average22, long22 = table_row(lines, out_dir, cp, rp, runs, **kwargs)
table.extend(average22)
long_table.extend(long22)

display(html_table([header]+average22))

Disjuncts-DRK-Disjuncts¶

%%capture
kwargs['context'] = 2
average23, long23 = table_row(lines, out_dir, cp, rp, runs, **kwargs)
table.extend(average23)
long_table.extend(long23)

display(html_table([header]+average23))

Disjuncts-ILE-Disjuncts¶

%%capture
kwargs['word_space'] = 'discrete'
kwargs['clustering'] = 'group'
average24, long24 = table_row(lines, out_dir, cp, rp, runs, **kwargs)
table.extend(average24)
long_table.extend(long24)

display(html_table([header]+average24))

All tests (all entries for multi-test `runs > (1.1)`)¶

display(html_table([header]+long_table))

print(UTC(), ':: finished, elapsed', str(round((time.time()-start)/60, 1)), 'min')
table_str = list2file(table, out_dir+'/short_table.txt')
if runs == (1,1):
    print('Results saved to', out_dir + '/short_table.txt')
else:
    long_table_str = list2file(long_table, out_dir+'/long_table.txt')
    print('Average results saved to', out_dir + '/short_table.txt\n'
          'Detailed results for every run saved to', out_dir + '/long_table.txt')

2018-10-03 17:35:20 UTC :: finished, elapsed 24.2 min
Results saved to /home/obaskov/py/language-learning/output/POC-English-NoAmb-2018-10-03/short_table.txt

Line	Corpus	Parsing	LW	RW	Gen.	Space	Rules	Silhouette	PA	PQ
14	POC-English-NoAmb	MST-fixed-manually	LW	.	none	cDRKc	10	0.45	64%	55%
15	POC-English-NoAmb	MST-fixed-manually	--	--	none	cDRKc	8	0.51	72%	64%
16	POC-English-NoAmb	LG-English	LW	.	none	cDRKc	13	0.52	68%	55%
17	POC-English-NoAmb	LG-English	--	--	none	cDRKc	11	0.58	75%	64%
18	POC-English-NoAmb	R=6-Weight=6:R-mst-weight=+1:R	LW	.	none	cDRKc	10	0.44	92%	56%
19	POC-English-NoAmb	R=6-Weight=6:R-mst-weight=+1:R	--	--	none	cDRKc	9	0.49	99%	65%
20	POC-English-NoAmb	R=6-Weight=1-mst-weight=+1:R	LW	.	none	cDRKc	24	0.16	90%	60%
21	POC-English-NoAmb	R=6-Weight=1-mst-weight=+1:R	--	--	none	cDRKc	22	0.18	90%	55%
22	POC-English-NoAmb	LG-ANY-mst-weight=+1:R	LW	.	none	cDRKc	10	0.44	98%	64%
23	POC-English-NoAmb	LG-ANY-mst-weight=+1:R	--	--	none	cDRKc	9	0.48	99%	65%

POC-English-NoAmb tests 2018-10-03¶