Gutenberg Children Books agglomerative clustering 2018-12-08

Agglomerative clustering, test_grammar updated 2018-10-19 , Link Grammar 5.5.1.
This notebook is shared as static Gutenberg-Children-Books-2018-12-08.html.
The "All tests" table is shared as 'all_tests_table.txt' in Gutenberg-Children-Books-2018-12-08 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
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)

Corpus test settings

corpus = 'Gutenberg-Children-Books-Caps'
dataset = 'LG-English'
kwargs = {
    'left_wall'     :   ''          ,
    'period'        :   False       ,
    'context'       :   1           ,
    'min_word_count':   31          ,   # ~7K words
    'word_space'    :   'sparse'    ,
    'clustering'    :   ['agglomerative', 'ward'],
    'clustering_metric' : ['silhouette', 'cosine'],
    'cluster_range' :   2000        ,   # 2000/1000/500/100/50/20
    'top_level'     :   0.01        ,
    'grammar_rules' :   2           ,
    'max_disjuncts' :   1000000     ,   # off
    'tmpath'        :   tmpath      , 
    'verbose'       :   'min'       ,
    'template_path' :   'poc-turtle',
    'linkage_limit' :   1000        }
rp = module_path + '/data/' + corpus + '/LG-E-clean/GCB-LG-English-clean.ull'
cp = rp  # corpus path = reference_path
runs = (1,1)
out_dir = module_path + '/output/' + corpus + '-' + str(UTC())[:10]

Tests: "LG English" corpus, words frequency > 30 occasions

display(html_table([header] + table))

Line	Corpus	Parsing	Space	Linkage	Affinity	G12n	Threshold	Rules	NN	SI	PA	PQ	F1	Top 5 cluster sizes
2.1	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	2000	---	0.0	47%	43%	0.50	[145, 138, 137, 126, 117]
2.2	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	1000	---	0.0	48%	43%	0.50	[384, 325, 286, 241, 236]
2.3	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	500	---	0.0	50%	44%	0.51	[1002, 466, 437, 384, 381]
2.4	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	100	---	0.0	60%	44%	0.46	[1414, 1094, 978, 824, 721]
2.5	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	50	---	0.0	62%	40%	0.42	[1414, 1350, 1146, 1116, 721]
2.6	Gutenberg-Children-Books-Caps	LG-English	cALWEd	ward	euclidean	none	---	20	---	0.0	62%	35%	0.36	[3281, 2466, 437, 371, 261]

%%capture
kwargs['cluster_range'] = 10
line = [[2.7, corpus, dataset, 0, 0, 'none']]
a, _, header, log, rules = wide_rows(line, out_dir, cp, rp, runs, **kwargs)
table.extend(a)

display(html_table([header] + a))

2018-12-10 08:57: still running...