ALE "Gutenberg Children Books" `2019-04-04` 50_clusters¶

"Gutenberg Children Books" corpus, new "LG-E-noQuotes" dataset (GC_LGEnglish_noQuotes_fullyParsed.ull),
trash filter off: min_word_count = 31,21,11,6,2,1, max_sentence_length off, Link Grammar 5.5.1.
Server 94, fresh clone of singnet repository, fresh ull environment ~ 2018-03-18 Check 2018-04-03 results for MWC=1

This notebook is shared as static cALEd-50-GCB-LG-E-noQuotes-2019-04-04.html.
Output data shared via cALEd-50-GCB-LG-E-noQuotes-2019-04-04 directory.

Basic settings¶

import os, sys, time
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, test_stats
from src.grammar_learner.read_files import check_dir, check_corpus
from src.grammar_learner.write_files import list2file
from src.grammar_learner.widgets import html_table
from src.grammar_learner.pqa_table import table_rows, params, wide_rows
tmpath = module_path + '/tmp/'
check_dir(tmpath, True, 'none')
start = time.time()
runs = (1,1)
print(UTC(), ':: module_path:', module_path)

2019-04-04 10:51:05 UTC :: module_path: /home/obaskov/94/language-learning

Corpus test settings¶

corpus = 'GCB' # 'Gutenberg-Children-Books' 
dataset = 'LG-E-noQuotes'  # 'LG-E-clean'
kwargs = {
    'left_wall'     :   ''          ,
    'period'        :   False       ,
    'context'       :   1           ,   # 1: connector-based word space
    'min_word_count':   1           ,   # 31,21,11,6,1
    'word_space'    :   'sparse'    ,
    'clustering'    :   ['agglomerative', 'ward'],
    'clustering_metric' : ['silhouette', 'cosine'],
    'cluster_range' :   [50]        ,
    'top_level'     :   0.01        ,
    'grammar_rules' :   2           ,   # disjunct-based grammar rules
    'max_disjuncts' :   1000000     ,   # off
    'stop_words'    :   []          ,
    'tmpath'        :   tmpath      ,
    'verbose'       :   'log+'      ,
    'template_path' :   'poc-turtle',
    'linkage_limit' :   1000        }
rp = module_path + '/data/GCB/LG-E-noQuotes/GC_LGEnglish_noQuotes_fullyParsed.ull'
cp = rp  # corpus path = reference_path
out_dir = module_path + '/output/' + 'cALEd-50-GCB-LG-E-noQuotes-' + str(UTC())[:10]
print(UTC(), '\n', out_dir)

2019-04-04 10:51:05 UTC 
 /home/obaskov/94/language-learning/output/cALEd-50-GCB-LG-E-noQuotes-2019-04-04

Tests: `min_word_count = 31, 21, 11, 6, 2, 1`¶

min_word_count = 31, 21, 11¶

%%capture
table = []
line = [['', corpus, dataset, 0, 0, 'none']]
kwargs['min_word_count'] = 31
a, _, header, log, rules = wide_rows(line, out_dir, cp, rp, runs, **kwargs)
header[0] = ''
table.extend(a)

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 3341 words, grammar learn time: 00:01:19, grammar test time: 00:52:40

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 4417 words, grammar learn time: 00:02:16, grammar test time: 01:03:53

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 6866 words, grammar learn time: 00:05:25, grammar test time: 01:12:58

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 10053 words, grammar learn time: 00:14:45, grammar test time: 01:24:00

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 19326 words, grammar learn time: 01:29:11, grammar test time: 01:28:00

display(html_table([header] + a)); print(test_stats(log))

Cleaned dictionary: 22641 words, grammar learn time: 02:13:52, grammar test time: 01:11:21

Save results¶

display(html_table([header] + table))

print(UTC(), ':: 6 tests finished, elapsed', str(round((time.time()-start)/3600.0, 1)), 'hours')
table_str = list2file(table, out_dir + '/all_tests_table.txt')
print('Results saved to', out_dir + '/all_tests_table.txt')

2019-04-04 22:10:53 UTC :: 6 tests finished, elapsed 11.3 hours
Results saved to /home/obaskov/94/language-learning/output/cALEd-50-GCB-LG-E-noQuotes-2019-04-04/all_tests_table.txt

Corpus	Parsing	Space	Linkage	Affinity	G12n	Threshold	Rules	MWC	NN	SI	PA	PQ	F1	Top 5 cluster sizes
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	31	---	0.0	80%	53%	0.59	[949, 738, 307, 240, 217]
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	21	---	0.0	82%	55%	0.61	[1309, 1225, 407, 364, 340]
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	11	---	0.0	85%	58%	0.63	[2447, 2256, 542, 525, 414]
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	6	---	0.0	88%	59%	0.64	[3686, 3002, 1408, 785, 397]
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	2	---	0.0	89%	61%	0.65	[14869, 1423, 706, 686, 616]
GCB	LG-E-noQuotes	cALWEd	ward	euclidean	none	---	50	1	---	0.0	89%	61%	0.65	[10050, 7610, 1683, 1370, 539]