lexsync 0.1.1__tar.gz → 0.2.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {lexsync-0.1.1 → lexsync-0.2.0}/.gitignore +9 -2
- {lexsync-0.1.1 → lexsync-0.2.0}/LICENSE +2 -1
- {lexsync-0.1.1 → lexsync-0.2.0}/LICENSE.note +34 -14
- {lexsync-0.1.1 → lexsync-0.2.0}/PKG-INFO +55 -47
- {lexsync-0.1.1 → lexsync-0.2.0}/README.md +52 -45
- {lexsync-0.1.1 → lexsync-0.2.0}/pyproject.toml +26 -24
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/__init__.py +5 -4
- lexsync-0.2.0/src/lexsync/blocks.py +194 -0
- lexsync-0.2.0/src/lexsync/cli.py +97 -0
- lexsync-0.2.0/src/lexsync/corpora.py +742 -0
- lexsync-0.2.0/src/lexsync/counterbalancing.py +933 -0
- lexsync-0.2.0/src/lexsync/data/en_example.csv +3001 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/data/registry.yaml +45 -17
- lexsync-0.2.0/src/lexsync/data/schema.yaml +429 -0
- lexsync-0.2.0/src/lexsync/datasheet.py +1926 -0
- lexsync-0.2.0/src/lexsync/design.py +581 -0
- lexsync-0.2.0/src/lexsync/generation.py +592 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/io_utils.py +254 -44
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/logging.py +30 -3
- lexsync-0.2.0/src/lexsync/matching.py +911 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/pairs.py +4 -2
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/paradigms.py +65 -9
- lexsync-0.2.0/src/lexsync/paths.py +74 -0
- lexsync-0.2.0/src/lexsync/querying.py +973 -0
- lexsync-0.2.0/src/lexsync/run_pipeline.py +841 -0
- lexsync-0.2.0/src/lexsync/scripting.py +2756 -0
- lexsync-0.2.0/src/lexsync/templates/jspsych/experiment_template.html +371 -0
- lexsync-0.2.0/src/lexsync/templates/opensesame/README.md +138 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/templates/opensesame/inline_send_triggers.py +6 -2
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/templates/opensesame/inline_trigger_setup.py +9 -11
- lexsync-0.2.0/src/lexsync/templates/psychopy/mock_port.py +24 -0
- lexsync-0.2.0/src/lexsync/templates/psychopy/trial_runner_template.py +892 -0
- lexsync-0.2.0/src/lexsync/validation.py +419 -0
- lexsync-0.2.0/tests/helper_corpus_files.py +74 -0
- lexsync-0.2.0/tests/test_apps.py +852 -0
- lexsync-0.2.0/tests/test_balance.py +294 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_blocks.py +44 -3
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_byte_parity.py +2 -2
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_categorisation.py +4 -3
- lexsync-0.2.0/tests/test_cli.py +173 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_config.py +13 -7
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_corpora.py +347 -9
- lexsync-0.2.0/tests/test_counterbalancing.py +271 -0
- lexsync-0.2.0/tests/test_datasheet.py +1198 -0
- lexsync-0.2.0/tests/test_datasheet_digests.py +92 -0
- lexsync-0.2.0/tests/test_demonstrations.py +186 -0
- lexsync-0.2.0/tests/test_derived_lexica.py +174 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_dimensions.py +67 -2
- lexsync-0.2.0/tests/test_docs_theme.py +27 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_exact_reductions.py +22 -0
- lexsync-0.2.0/tests/test_exclude_words.py +280 -0
- lexsync-0.2.0/tests/test_generation_accents.py +148 -0
- lexsync-0.2.0/tests/test_generation_exclude.py +235 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_hash_primitive.py +17 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_injection.py +114 -7
- lexsync-0.2.0/tests/test_install_lines.py +139 -0
- lexsync-0.2.0/tests/test_instructions.py +482 -0
- lexsync-0.2.0/tests/test_io_utils.py +474 -0
- lexsync-0.2.0/tests/test_jspsych_export.py +393 -0
- lexsync-0.2.0/tests/test_known_forms_build.py +70 -0
- lexsync-0.2.0/tests/test_lexicon_provenance.py +191 -0
- lexsync-0.2.0/tests/test_masked_priming.py +217 -0
- lexsync-0.2.0/tests/test_matching.py +1172 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_metadata.py +136 -24
- lexsync-0.2.0/tests/test_normalise.py +172 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_norms.py +154 -2
- lexsync-0.2.0/tests/test_osexp_validator.py +1292 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_pairs.py +22 -1
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_parity.py +5 -2
- lexsync-0.2.0/tests/test_psychopy_export.py +1132 -0
- lexsync-0.2.0/tests/test_querying.py +526 -0
- lexsync-0.2.0/tests/test_run_pipeline.py +1444 -0
- lexsync-0.2.0/tests/test_scripting.py +777 -0
- lexsync-0.2.0/tests/test_templates.py +190 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_timing.py +57 -1
- lexsync-0.2.0/tests/test_trial_order.py +252 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_trial_timing.py +74 -0
- lexsync-0.2.0/tests/test_twin_edge.py +203 -0
- lexsync-0.2.0/tests/test_validation.py +372 -0
- lexsync-0.1.1/src/lexsync/blocks.py +0 -149
- lexsync-0.1.1/src/lexsync/cli.py +0 -85
- lexsync-0.1.1/src/lexsync/corpora.py +0 -331
- lexsync-0.1.1/src/lexsync/counterbalancing.py +0 -399
- lexsync-0.1.1/src/lexsync/data/en_example.csv +0 -3001
- lexsync-0.1.1/src/lexsync/data/schema.yaml +0 -193
- lexsync-0.1.1/src/lexsync/datasheet.py +0 -752
- lexsync-0.1.1/src/lexsync/generation.py +0 -280
- lexsync-0.1.1/src/lexsync/matching.py +0 -614
- lexsync-0.1.1/src/lexsync/querying.py +0 -481
- lexsync-0.1.1/src/lexsync/run_pipeline.py +0 -425
- lexsync-0.1.1/src/lexsync/scripting.py +0 -845
- lexsync-0.1.1/src/lexsync/templates/jspsych/experiment_template.html +0 -223
- lexsync-0.1.1/src/lexsync/templates/opensesame/README.md +0 -36
- lexsync-0.1.1/src/lexsync/templates/psychopy/mock_port.py +0 -23
- lexsync-0.1.1/src/lexsync/templates/psychopy/trial_runner_template.py +0 -300
- lexsync-0.1.1/src/lexsync/validation.py +0 -267
- lexsync-0.1.1/tests/test_apps.py +0 -194
- lexsync-0.1.1/tests/test_balance.py +0 -155
- lexsync-0.1.1/tests/test_cli.py +0 -76
- lexsync-0.1.1/tests/test_counterbalancing.py +0 -81
- lexsync-0.1.1/tests/test_datasheet.py +0 -435
- lexsync-0.1.1/tests/test_io_utils.py +0 -104
- lexsync-0.1.1/tests/test_jspsych_export.py +0 -62
- lexsync-0.1.1/tests/test_matching.py +0 -501
- lexsync-0.1.1/tests/test_osexp_validator.py +0 -50
- lexsync-0.1.1/tests/test_psychopy_export.py +0 -125
- lexsync-0.1.1/tests/test_querying.py +0 -268
- lexsync-0.1.1/tests/test_run_pipeline.py +0 -286
- lexsync-0.1.1/tests/test_scripting.py +0 -214
- lexsync-0.1.1/tests/test_templates.py +0 -71
- lexsync-0.1.1/tests/test_validation.py +0 -185
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/__main__.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/data/es_example.csv +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/data/zh_example.csv +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/src/lexsync/py.typed +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/conftest.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/helper_repo.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_es_gender_build.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_generation_subsyllabic.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_init.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_paradigms.py +0 -0
- {lexsync-0.1.1 → lexsync-0.2.0}/tests/test_readme.py +0 -0
|
@@ -49,6 +49,12 @@ R_workflow/docs/
|
|
|
49
49
|
python_workflow/site/
|
|
50
50
|
/site/
|
|
51
51
|
|
|
52
|
+
# --- Participant data -------------------------------------------------------
|
|
53
|
+
# A generated PsychoPy script writes each run's data beside itself, under data/,
|
|
54
|
+
# so a demonstration run in output/experiments/ would otherwise leave participant
|
|
55
|
+
# data where a commit could pick it up.
|
|
56
|
+
output/experiments/data/
|
|
57
|
+
|
|
52
58
|
# --- session build/check temporaries ---------------------------------------
|
|
53
59
|
/.rcheck_tmp/
|
|
54
60
|
/.regen_R.log
|
|
@@ -56,8 +62,9 @@ python_workflow/site/
|
|
|
56
62
|
/.audit_findings.txt
|
|
57
63
|
|
|
58
64
|
# --- Credentials -----------------------------------------------------------
|
|
59
|
-
#
|
|
60
|
-
# Twine reads an upload token from .pypirc in
|
|
65
|
+
# publish.yml uploads to PyPI by trusted publishing and needs no token, but a
|
|
66
|
+
# manual upload with twine does. Twine reads an upload token from .pypirc in
|
|
67
|
+
# plaintext. It looks in the home
|
|
61
68
|
# directory by default, but --config-file will take one from anywhere, and a copy
|
|
62
69
|
# made for a single upload is the kind that gets left behind. Nothing else listed
|
|
63
70
|
# here is expected to exist. Those patterns are defensive: the repository is
|
|
@@ -24,7 +24,8 @@ SOFTWARE.
|
|
|
24
24
|
The MIT License above covers the lexsync source code. The lexical corpora that
|
|
25
25
|
the package bundles or fetches are NOT covered by it. Three of the data files
|
|
26
26
|
shipped inside the package, lexsync/data/en_example.csv, es_example.csv and
|
|
27
|
-
zh_example.csv, are wordfreq derivatives distributed under CC BY-SA 4.0
|
|
27
|
+
zh_example.csv, are wordfreq derivatives distributed under CC BY-SA 4.0, the
|
|
28
|
+
first of them screened against SUBTLEX-US, which is under the same licence. Their
|
|
28
29
|
terms are set out in full in LICENSE.note, which ships beside this file. The
|
|
29
30
|
repository keeps the fuller record, with the retrieval date and checksum of
|
|
30
31
|
every derived file, at
|
|
@@ -28,6 +28,17 @@ must give appropriate credit to the corpus authors, indicate that changes were
|
|
|
28
28
|
made, and place any adaptation under the same licence. Everything else in the
|
|
29
29
|
package, including the other files under lexsync/data, is MIT.
|
|
30
30
|
|
|
31
|
+
One of the three has a second direct source. en_example.csv holds 3,000 of the
|
|
32
|
+
forms of the repository's screened English lexicon, corpora/derived/en_v2.csv.
|
|
33
|
+
That lexicon holds every form of corpora/derived/en.csv and marks as nonstandard
|
|
34
|
+
each form SUBTLEX-US (Brysbaert & New, 2009) lacks or whose uses there begin with
|
|
35
|
+
a lower-case letter less than a tenth of the time, and en_example.csv holds only
|
|
36
|
+
forms it leaves unmarked. SUBTLEX-US, which openlexicon distributes under the same
|
|
37
|
+
licence, CC BY-SA 4.0, thus chose the file's rows, and the file's N and OLD20 count
|
|
38
|
+
neighbours among the forms en_v2.csv leaves unmarked.
|
|
39
|
+
SUBTLEX-US is therefore a direct source of en_example.csv as well as one of the
|
|
40
|
+
lists wordfreq blends.
|
|
41
|
+
|
|
31
42
|
Two further notices belong with these three files, because CC BY-SA 4.0 asks
|
|
32
43
|
that they be kept with the material. The copyright notice carried by wordfreq's
|
|
33
44
|
own licence file is 'Copyright 2022 Robyn Speer', and it is retained here. As
|
|
@@ -50,7 +61,8 @@ package reads, and two orthographic-neighbourhood variables computed here from
|
|
|
50
61
|
the word list itself: Coltheart's N, the count of same-length neighbours
|
|
51
62
|
reachable by a single letter substitution, and OLD20, the mean Levenshtein
|
|
52
63
|
distance to the twenty nearest words. Each row names its origin in the 'source'
|
|
53
|
-
column.
|
|
64
|
+
column. For en_example.csv, both variables count neighbours among the forms the
|
|
65
|
+
SUBTLEX-US screen keeps, as described above.
|
|
54
66
|
|
|
55
67
|
|
|
56
68
|
Credit for SUBTLEX
|
|
@@ -64,35 +76,43 @@ from it: work derived from wordfreq must credit the SUBTLEX authors, and it must
|
|
|
64
76
|
remain clear that SUBTLEX is freely available data. Both are honoured here. The
|
|
65
77
|
lists are freely available, and their authors are credited as follows.
|
|
66
78
|
|
|
67
|
-
Brysbaert, M., & New, B. (2009).
|
|
68
|
-
|
|
79
|
+
Brysbaert, M., & New, B. (2009). Moving beyond Kučera and Francis: A critical
|
|
80
|
+
evaluation of current word frequency norms and the introduction of a new and
|
|
81
|
+
improved word frequency measure for American English. Behavior Research
|
|
82
|
+
Methods, 41(4), 977-990. https://doi.org/10.3758/BRM.41.4.977
|
|
69
83
|
|
|
70
84
|
van Heuven, W. J. B., Mandera, P., Keuleers, E., & Brysbaert, M. (2014).
|
|
71
|
-
SUBTLEX-UK
|
|
85
|
+
SUBTLEX-UK: A new and improved word frequency database for British English.
|
|
86
|
+
Quarterly Journal of Experimental Psychology, 67(6), 1176-1190.
|
|
72
87
|
https://doi.org/10.1080/17470218.2013.850521
|
|
73
88
|
|
|
74
|
-
Cai, Q., & Brysbaert, M. (2010). SUBTLEX-CH
|
|
89
|
+
Cai, Q., & Brysbaert, M. (2010). SUBTLEX-CH: Chinese word and character
|
|
90
|
+
frequencies based on film subtitles. PLoS ONE, 5(6), e10729.
|
|
75
91
|
https://doi.org/10.1371/journal.pone.0010729
|
|
76
92
|
|
|
77
93
|
Brysbaert, M., Buchmeier, M., Conrad, M., Jacobs, A. M., Bölte, J., & Böhl, A.
|
|
78
|
-
(2011).
|
|
79
|
-
|
|
94
|
+
(2011). The word frequency effect: A review of recent developments and
|
|
95
|
+
implications for the choice of frequency estimates in German. Experimental
|
|
96
|
+
Psychology, 58(5), 412-424. https://doi.org/10.1027/1618-3169/a000123
|
|
80
97
|
|
|
81
|
-
Keuleers, E., Brysbaert, M., & New, B. (2010). SUBTLEX-NL
|
|
82
|
-
|
|
98
|
+
Keuleers, E., Brysbaert, M., & New, B. (2010). SUBTLEX-NL: A new measure for
|
|
99
|
+
Dutch word frequency based on film subtitles. Behavior Research Methods,
|
|
100
|
+
42(3), 643-650. https://doi.org/10.3758/BRM.42.3.643
|
|
83
101
|
|
|
84
102
|
SUBTLEX contributes to the blend rather than constituting it. No file bundled
|
|
85
|
-
here is a copy of a SUBTLEX list
|
|
86
|
-
|
|
87
|
-
anything shipped in
|
|
103
|
+
here is a copy of a SUBTLEX list. As described above, SUBTLEX-US screened the rows
|
|
104
|
+
of en_example.csv. Otherwise, the SUBTLEX resources the corpus registry offers are
|
|
105
|
+
separate corpora a user may fetch directly, not the source of anything shipped in
|
|
106
|
+
this package.
|
|
88
107
|
|
|
89
108
|
|
|
90
109
|
Corpora fetched at run time
|
|
91
110
|
---------------------------
|
|
92
111
|
|
|
93
112
|
Corpora downloaded on demand through the registry are not redistributed with
|
|
94
|
-
this package. Each carries its own licence and citation, recorded in
|
|
95
|
-
registry
|
|
113
|
+
this package. Each carries its own licence and citation, recorded in its entry
|
|
114
|
+
in lexsync/data/registry.yaml. list_corpora() reports the citation, and
|
|
115
|
+
fetch_corpus() prints it after a download. The optional wordfreq
|
|
96
116
|
connector produces lexica under the same terms as the bundled derivatives,
|
|
97
117
|
since they come from the same source.
|
|
98
118
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.5
|
|
2
2
|
Name: lexsync
|
|
3
|
-
Version: 0.
|
|
3
|
+
Version: 0.2.0
|
|
4
4
|
Summary: Lexical optimisation and hardware-timed experiment generation
|
|
5
5
|
Project-URL: Homepage, https://github.com/pablobernabeu/lexsync
|
|
6
6
|
Project-URL: Repository, https://github.com/pablobernabeu/lexsync
|
|
@@ -23,6 +23,7 @@ Classifier: Programming Language :: Python :: 3.11
|
|
|
23
23
|
Classifier: Programming Language :: Python :: 3.12
|
|
24
24
|
Classifier: Programming Language :: Python :: 3.13
|
|
25
25
|
Classifier: Programming Language :: Python :: 3.14
|
|
26
|
+
Classifier: Programming Language :: Python :: 3.15
|
|
26
27
|
Classifier: Topic :: Scientific/Engineering
|
|
27
28
|
Classifier: Topic :: Scientific/Engineering :: Information Analysis
|
|
28
29
|
Classifier: Typing :: Typed
|
|
@@ -37,7 +38,7 @@ Requires-Dist: wordfreq<4,>=3.0; extra == 'corpora'
|
|
|
37
38
|
Provides-Extra: dev
|
|
38
39
|
Requires-Dist: build; extra == 'dev'
|
|
39
40
|
Requires-Dist: pytest>=7.0; extra == 'dev'
|
|
40
|
-
Requires-Dist: streamlit>=1.
|
|
41
|
+
Requires-Dist: streamlit>=1.49; extra == 'dev'
|
|
41
42
|
Requires-Dist: twine; extra == 'dev'
|
|
42
43
|
Provides-Extra: experiment
|
|
43
44
|
Requires-Dist: psychopy>=2024.1; extra == 'experiment'
|
|
@@ -48,11 +49,11 @@ Description-Content-Type: text/markdown
|
|
|
48
49
|
|
|
49
50
|
<!-- badges: start -->
|
|
50
51
|
[](https://pypi.org/project/lexsync/)
|
|
51
|
-
[](https://doi.org/10.5281/zenodo.22906962)
|
|
52
52
|
[](https://github.com/pablobernabeu/lexsync/actions/workflows/python-tests.yaml)
|
|
53
53
|
[](https://lifecycle.r-lib.org/articles/stages.html#experimental)
|
|
54
54
|
[](https://opensource.org/license/MIT)
|
|
55
55
|
[](https://creativecommons.org/licenses/by-sa/4.0/)
|
|
56
|
+
[](https://doi.org/10.5281/zenodo.22906962)
|
|
56
57
|
<!-- badges: end -->
|
|
57
58
|
|
|
58
59
|
Lexical optimisation and hardware-timed experiment generation.
|
|
@@ -65,12 +66,11 @@ carry hardware triggers injected at stimulus onset for EEG/ERP synchronisation,
|
|
|
65
66
|
and the browser target is a single shareable HTML file.
|
|
66
67
|
|
|
67
68
|
This is the feature-parity twin of [the R
|
|
68
|
-
package](https://pablobernabeu.github.io/lexsync/r/) of the same name
|
|
69
|
-
|
|
70
|
-
under the deterministic matching methods
|
|
71
|
-
released under one version.
|
|
69
|
+
package](https://pablobernabeu.github.io/lexsync/r/) of the same name. The two
|
|
70
|
+
are built from one repository and released under one version, and their engines
|
|
71
|
+
select byte-identical stimuli under the deterministic matching methods.
|
|
72
72
|
|
|
73
|
-
Documentation, including the guides and the
|
|
73
|
+
Documentation, including the guides and the API reference, is at
|
|
74
74
|
<https://pablobernabeu.github.io/lexsync/python/>.
|
|
75
75
|
|
|
76
76
|
## Install
|
|
@@ -81,13 +81,13 @@ The package is on [PyPI](https://pypi.org/project/lexsync/):
|
|
|
81
81
|
pip install lexsync
|
|
82
82
|
```
|
|
83
83
|
|
|
84
|
-
The `corpora` extra adds the 'wordfreq' connector, which
|
|
85
|
-
languages through a single dependency, and the
|
|
86
|
-
'PsychoPy' and 'pyserial'.
|
|
87
|
-
a generated experiment on hardware,
|
|
88
|
-
supports Python 3.10 to 3.12
|
|
89
|
-
message naming psychopy, while the rest of the
|
|
90
|
-
version. [Getting
|
|
84
|
+
The `corpora` extra adds the 'wordfreq' connector, which builds lexica for thirty
|
|
85
|
+
registered languages ('wordfreq' covers 42) through a single dependency, and the
|
|
86
|
+
`experiment` extra adds 'PsychoPy' and 'pyserial'. The `experiment` extra is
|
|
87
|
+
needed only to run a generated experiment on hardware, not to generate one.
|
|
88
|
+
'PsychoPy' currently supports Python 3.10 to 3.12. On a newer interpreter,
|
|
89
|
+
installing that extra fails with a message naming psychopy, while the rest of the
|
|
90
|
+
package installs on any supported version. [Getting
|
|
91
91
|
started](https://pablobernabeu.github.io/lexsync/python/getting-started/)
|
|
92
92
|
explains why and suggests an interpreter for hardware work. An extra goes in
|
|
93
93
|
brackets after the package name:
|
|
@@ -145,13 +145,13 @@ report = lexsync.match_report(
|
|
|
145
145
|
print(report["comparisons"].to_string(index=False))
|
|
146
146
|
```
|
|
147
147
|
|
|
148
|
-
The report
|
|
149
|
-
|
|
150
|
-
|
|
151
|
-
|
|
152
|
-
|
|
148
|
+
The report measures what the matching achieved. Frequency, the manipulation,
|
|
149
|
+
separates the conditions by nearly five standard deviations. Each control
|
|
150
|
+
dimension passes a two one-sided tests procedure against a bound of *d* = 0.5, so
|
|
151
|
+
it is shown to be equivalent rather than merely failing to differ significantly.
|
|
152
|
+
The [Matching and designs
|
|
153
153
|
guide](https://pablobernabeu.github.io/lexsync/python/matching-and-designs/)
|
|
154
|
-
|
|
154
|
+
explains the report column by column.
|
|
155
155
|
|
|
156
156
|
## Use
|
|
157
157
|
|
|
@@ -170,9 +170,13 @@ cd lexsync
|
|
|
170
170
|
import yaml, lexsync
|
|
171
171
|
|
|
172
172
|
schema = yaml.safe_load(open("config/schema.yaml"))
|
|
173
|
-
lex = lexsync.load_lexicon("corpora/derived/en.csv", schema, "english")
|
|
174
|
-
|
|
175
173
|
design = yaml.safe_load(open("config/design_en_freqcontrast.yaml"))
|
|
174
|
+
lex = lexsync.load_lexicon(design["lexicon"], schema, "english") # en_v2.csv
|
|
175
|
+
|
|
176
|
+
# The design leaves out the words its exclude_words file lists, as `lexsync run` does.
|
|
177
|
+
with open(design["exclude_words"]["path"], encoding="utf-8") as handle:
|
|
178
|
+
excluded = {w.strip() for w in handle if w.strip() and not w.startswith("#")}
|
|
179
|
+
lex = lex[~lex["word"].isin(excluded)]
|
|
176
180
|
pool = lexsync.build_pool(lex, design["pool_filters"])
|
|
177
181
|
stim = lexsync.match_stimuli(pool, design, schema)
|
|
178
182
|
# Not optional: this assigns the counterbalancing lists and draws the trial
|
|
@@ -181,14 +185,12 @@ stim = lexsync.match_stimuli(pool, design, schema)
|
|
|
181
185
|
stim = lexsync.counterbalance(stim, design, schema)
|
|
182
186
|
|
|
183
187
|
report = lexsync.match_report(stim, ["length", "frequency", "n_density", "old20"], schema)
|
|
184
|
-
lexsync.export_experiments(
|
|
185
|
-
lexsync.scripting.assign_triggers(stim), design, schema, "output/experiments"
|
|
186
|
-
)
|
|
188
|
+
lexsync.export_experiments(stim, design, schema, "output/experiments")
|
|
187
189
|
```
|
|
188
190
|
|
|
189
|
-
The same operations are available from the command line
|
|
190
|
-
|
|
191
|
-
corpora and
|
|
191
|
+
The same operations are available from the command line. The first command below
|
|
192
|
+
runs a whole design, again from the root of the clone, and the other two list the
|
|
193
|
+
registered corpora and derive a new lexicon:
|
|
192
194
|
|
|
193
195
|
```bash
|
|
194
196
|
lexsync run config/design_en_freqcontrast.yaml
|
|
@@ -196,17 +198,24 @@ lexsync corpora list
|
|
|
196
198
|
lexsync fetch fr # build a French lexicon via wordfreq
|
|
197
199
|
```
|
|
198
200
|
|
|
201
|
+
Outside a clone, name the schema with `--schema` or leave it out. When it is left
|
|
202
|
+
out, `lexsync run` looks for `schema.yaml` beside the designs and falls back to the
|
|
203
|
+
copy bundled with the package. It reports on standard error which file it used. A
|
|
204
|
+
command that fails prints lexsync's message and exits with status 1, and
|
|
205
|
+
`lexsync --traceback run ...` shows the Python traceback as well.
|
|
206
|
+
|
|
199
207
|
## Citation
|
|
200
208
|
|
|
201
209
|
Cite the software. The authoritative record is
|
|
202
210
|
[`CITATION.cff`](https://github.com/pablobernabeu/lexsync/blob/main/CITATION.cff)
|
|
203
211
|
at the root of the repository, which GitHub renders into several formats through
|
|
204
|
-
its 'Cite this repository' button.
|
|
205
|
-
|
|
206
|
-
|
|
207
|
-
Zenodo archive,
|
|
208
|
-
`CITATION.cff`
|
|
209
|
-
|
|
212
|
+
its 'Cite this repository' button. PyPI assigns no DOI, so the formatted
|
|
213
|
+
reference for this package on the [About
|
|
214
|
+
page](https://pablobernabeu.github.io/lexsync/python/about/) gives its version and
|
|
215
|
+
the concept DOI of the Zenodo archive, <https://doi.org/10.5281/zenodo.22906962>.
|
|
216
|
+
`CITATION.cff` describes both packages, so it gives the same DOI and lists each
|
|
217
|
+
package's identifiers, among them the DOI that CRAN assigned to the R package. A
|
|
218
|
+
manuscript describing lexsync is in preparation.
|
|
210
219
|
|
|
211
220
|
Cite the corpus as well as the software. The corpora are third-party work with
|
|
212
221
|
their own terms, and each is credited, with its licence and retrieval date, in
|
|
@@ -214,21 +223,20 @@ their own terms, and each is credited, with its licence and retrieval date, in
|
|
|
214
223
|
|
|
215
224
|
## Licence
|
|
216
225
|
|
|
217
|
-
MIT for the code. The bundled corpus derivatives are
|
|
218
|
-
three example lexica
|
|
219
|
-
`lexsync/data/
|
|
220
|
-
'wordfreq' and
|
|
221
|
-
|
|
222
|
-
keep the same terms on any
|
|
223
|
-
|
|
224
|
-
|
|
225
|
-
and checksum of every derived file, in
|
|
226
|
+
MIT for the code. The bundled corpus derivatives are released under CC BY-SA 4.0
|
|
227
|
+
instead. They are the three example lexica installed as
|
|
228
|
+
`lexsync/data/en_example.csv`, `lexsync/data/es_example.csv` and
|
|
229
|
+
`lexsync/data/zh_example.csv`, derived from 'wordfreq' and, for English, screened
|
|
230
|
+
against SUBTLEX-US. CC BY-SA 4.0 asks anyone who redistributes them to credit the
|
|
231
|
+
corpus authors, say that changes were made and keep the same terms on any
|
|
232
|
+
adaptation. `LICENSE.note`, installed alongside `LICENSE`, states these terms
|
|
233
|
+
inside the distribution. The repository keeps the fuller record, including the
|
|
234
|
+
retrieval date and checksum of every derived file, in
|
|
226
235
|
[`LICENSE-DATA`](https://github.com/pablobernabeu/lexsync/blob/main/LICENSE-DATA).
|
|
227
236
|
|
|
228
237
|
## Contributing
|
|
229
238
|
|
|
230
239
|
Issues and pull requests are welcome, on the [issue
|
|
231
240
|
tracker](https://github.com/pablobernabeu/lexsync/issues) of the repository that
|
|
232
|
-
holds both twins. A report that includes the design YAML and the run log
|
|
233
|
-
|
|
234
|
-
ran.
|
|
241
|
+
holds both twins. A report that includes the design YAML and the run log can be
|
|
242
|
+
acted on, because together they record the inputs and every step that ran.
|
|
@@ -2,11 +2,11 @@
|
|
|
2
2
|
|
|
3
3
|
<!-- badges: start -->
|
|
4
4
|
[](https://pypi.org/project/lexsync/)
|
|
5
|
-
[](https://doi.org/10.5281/zenodo.22906962)
|
|
6
5
|
[](https://github.com/pablobernabeu/lexsync/actions/workflows/python-tests.yaml)
|
|
7
6
|
[](https://lifecycle.r-lib.org/articles/stages.html#experimental)
|
|
8
7
|
[](https://opensource.org/license/MIT)
|
|
9
8
|
[](https://creativecommons.org/licenses/by-sa/4.0/)
|
|
9
|
+
[](https://doi.org/10.5281/zenodo.22906962)
|
|
10
10
|
<!-- badges: end -->
|
|
11
11
|
|
|
12
12
|
Lexical optimisation and hardware-timed experiment generation.
|
|
@@ -19,12 +19,11 @@ carry hardware triggers injected at stimulus onset for EEG/ERP synchronisation,
|
|
|
19
19
|
and the browser target is a single shareable HTML file.
|
|
20
20
|
|
|
21
21
|
This is the feature-parity twin of [the R
|
|
22
|
-
package](https://pablobernabeu.github.io/lexsync/r/) of the same name
|
|
23
|
-
|
|
24
|
-
under the deterministic matching methods
|
|
25
|
-
released under one version.
|
|
22
|
+
package](https://pablobernabeu.github.io/lexsync/r/) of the same name. The two
|
|
23
|
+
are built from one repository and released under one version, and their engines
|
|
24
|
+
select byte-identical stimuli under the deterministic matching methods.
|
|
26
25
|
|
|
27
|
-
Documentation, including the guides and the
|
|
26
|
+
Documentation, including the guides and the API reference, is at
|
|
28
27
|
<https://pablobernabeu.github.io/lexsync/python/>.
|
|
29
28
|
|
|
30
29
|
## Install
|
|
@@ -35,13 +34,13 @@ The package is on [PyPI](https://pypi.org/project/lexsync/):
|
|
|
35
34
|
pip install lexsync
|
|
36
35
|
```
|
|
37
36
|
|
|
38
|
-
The `corpora` extra adds the 'wordfreq' connector, which
|
|
39
|
-
languages through a single dependency, and the
|
|
40
|
-
'PsychoPy' and 'pyserial'.
|
|
41
|
-
a generated experiment on hardware,
|
|
42
|
-
supports Python 3.10 to 3.12
|
|
43
|
-
message naming psychopy, while the rest of the
|
|
44
|
-
version. [Getting
|
|
37
|
+
The `corpora` extra adds the 'wordfreq' connector, which builds lexica for thirty
|
|
38
|
+
registered languages ('wordfreq' covers 42) through a single dependency, and the
|
|
39
|
+
`experiment` extra adds 'PsychoPy' and 'pyserial'. The `experiment` extra is
|
|
40
|
+
needed only to run a generated experiment on hardware, not to generate one.
|
|
41
|
+
'PsychoPy' currently supports Python 3.10 to 3.12. On a newer interpreter,
|
|
42
|
+
installing that extra fails with a message naming psychopy, while the rest of the
|
|
43
|
+
package installs on any supported version. [Getting
|
|
45
44
|
started](https://pablobernabeu.github.io/lexsync/python/getting-started/)
|
|
46
45
|
explains why and suggests an interpreter for hardware work. An extra goes in
|
|
47
46
|
brackets after the package name:
|
|
@@ -99,13 +98,13 @@ report = lexsync.match_report(
|
|
|
99
98
|
print(report["comparisons"].to_string(index=False))
|
|
100
99
|
```
|
|
101
100
|
|
|
102
|
-
The report
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
101
|
+
The report measures what the matching achieved. Frequency, the manipulation,
|
|
102
|
+
separates the conditions by nearly five standard deviations. Each control
|
|
103
|
+
dimension passes a two one-sided tests procedure against a bound of *d* = 0.5, so
|
|
104
|
+
it is shown to be equivalent rather than merely failing to differ significantly.
|
|
105
|
+
The [Matching and designs
|
|
107
106
|
guide](https://pablobernabeu.github.io/lexsync/python/matching-and-designs/)
|
|
108
|
-
|
|
107
|
+
explains the report column by column.
|
|
109
108
|
|
|
110
109
|
## Use
|
|
111
110
|
|
|
@@ -124,9 +123,13 @@ cd lexsync
|
|
|
124
123
|
import yaml, lexsync
|
|
125
124
|
|
|
126
125
|
schema = yaml.safe_load(open("config/schema.yaml"))
|
|
127
|
-
lex = lexsync.load_lexicon("corpora/derived/en.csv", schema, "english")
|
|
128
|
-
|
|
129
126
|
design = yaml.safe_load(open("config/design_en_freqcontrast.yaml"))
|
|
127
|
+
lex = lexsync.load_lexicon(design["lexicon"], schema, "english") # en_v2.csv
|
|
128
|
+
|
|
129
|
+
# The design leaves out the words its exclude_words file lists, as `lexsync run` does.
|
|
130
|
+
with open(design["exclude_words"]["path"], encoding="utf-8") as handle:
|
|
131
|
+
excluded = {w.strip() for w in handle if w.strip() and not w.startswith("#")}
|
|
132
|
+
lex = lex[~lex["word"].isin(excluded)]
|
|
130
133
|
pool = lexsync.build_pool(lex, design["pool_filters"])
|
|
131
134
|
stim = lexsync.match_stimuli(pool, design, schema)
|
|
132
135
|
# Not optional: this assigns the counterbalancing lists and draws the trial
|
|
@@ -135,14 +138,12 @@ stim = lexsync.match_stimuli(pool, design, schema)
|
|
|
135
138
|
stim = lexsync.counterbalance(stim, design, schema)
|
|
136
139
|
|
|
137
140
|
report = lexsync.match_report(stim, ["length", "frequency", "n_density", "old20"], schema)
|
|
138
|
-
lexsync.export_experiments(
|
|
139
|
-
lexsync.scripting.assign_triggers(stim), design, schema, "output/experiments"
|
|
140
|
-
)
|
|
141
|
+
lexsync.export_experiments(stim, design, schema, "output/experiments")
|
|
141
142
|
```
|
|
142
143
|
|
|
143
|
-
The same operations are available from the command line
|
|
144
|
-
|
|
145
|
-
corpora and
|
|
144
|
+
The same operations are available from the command line. The first command below
|
|
145
|
+
runs a whole design, again from the root of the clone, and the other two list the
|
|
146
|
+
registered corpora and derive a new lexicon:
|
|
146
147
|
|
|
147
148
|
```bash
|
|
148
149
|
lexsync run config/design_en_freqcontrast.yaml
|
|
@@ -150,17 +151,24 @@ lexsync corpora list
|
|
|
150
151
|
lexsync fetch fr # build a French lexicon via wordfreq
|
|
151
152
|
```
|
|
152
153
|
|
|
154
|
+
Outside a clone, name the schema with `--schema` or leave it out. When it is left
|
|
155
|
+
out, `lexsync run` looks for `schema.yaml` beside the designs and falls back to the
|
|
156
|
+
copy bundled with the package. It reports on standard error which file it used. A
|
|
157
|
+
command that fails prints lexsync's message and exits with status 1, and
|
|
158
|
+
`lexsync --traceback run ...` shows the Python traceback as well.
|
|
159
|
+
|
|
153
160
|
## Citation
|
|
154
161
|
|
|
155
162
|
Cite the software. The authoritative record is
|
|
156
163
|
[`CITATION.cff`](https://github.com/pablobernabeu/lexsync/blob/main/CITATION.cff)
|
|
157
164
|
at the root of the repository, which GitHub renders into several formats through
|
|
158
|
-
its 'Cite this repository' button.
|
|
159
|
-
|
|
160
|
-
|
|
161
|
-
Zenodo archive,
|
|
162
|
-
`CITATION.cff`
|
|
163
|
-
|
|
165
|
+
its 'Cite this repository' button. PyPI assigns no DOI, so the formatted
|
|
166
|
+
reference for this package on the [About
|
|
167
|
+
page](https://pablobernabeu.github.io/lexsync/python/about/) gives its version and
|
|
168
|
+
the concept DOI of the Zenodo archive, <https://doi.org/10.5281/zenodo.22906962>.
|
|
169
|
+
`CITATION.cff` describes both packages, so it gives the same DOI and lists each
|
|
170
|
+
package's identifiers, among them the DOI that CRAN assigned to the R package. A
|
|
171
|
+
manuscript describing lexsync is in preparation.
|
|
164
172
|
|
|
165
173
|
Cite the corpus as well as the software. The corpora are third-party work with
|
|
166
174
|
their own terms, and each is credited, with its licence and retrieval date, in
|
|
@@ -168,21 +176,20 @@ their own terms, and each is credited, with its licence and retrieval date, in
|
|
|
168
176
|
|
|
169
177
|
## Licence
|
|
170
178
|
|
|
171
|
-
MIT for the code. The bundled corpus derivatives are
|
|
172
|
-
three example lexica
|
|
173
|
-
`lexsync/data/
|
|
174
|
-
'wordfreq' and
|
|
175
|
-
|
|
176
|
-
keep the same terms on any
|
|
177
|
-
|
|
178
|
-
|
|
179
|
-
and checksum of every derived file, in
|
|
179
|
+
MIT for the code. The bundled corpus derivatives are released under CC BY-SA 4.0
|
|
180
|
+
instead. They are the three example lexica installed as
|
|
181
|
+
`lexsync/data/en_example.csv`, `lexsync/data/es_example.csv` and
|
|
182
|
+
`lexsync/data/zh_example.csv`, derived from 'wordfreq' and, for English, screened
|
|
183
|
+
against SUBTLEX-US. CC BY-SA 4.0 asks anyone who redistributes them to credit the
|
|
184
|
+
corpus authors, say that changes were made and keep the same terms on any
|
|
185
|
+
adaptation. `LICENSE.note`, installed alongside `LICENSE`, states these terms
|
|
186
|
+
inside the distribution. The repository keeps the fuller record, including the
|
|
187
|
+
retrieval date and checksum of every derived file, in
|
|
180
188
|
[`LICENSE-DATA`](https://github.com/pablobernabeu/lexsync/blob/main/LICENSE-DATA).
|
|
181
189
|
|
|
182
190
|
## Contributing
|
|
183
191
|
|
|
184
192
|
Issues and pull requests are welcome, on the [issue
|
|
185
193
|
tracker](https://github.com/pablobernabeu/lexsync/issues) of the repository that
|
|
186
|
-
holds both twins. A report that includes the design YAML and the run log
|
|
187
|
-
|
|
188
|
-
ran.
|
|
194
|
+
holds both twins. A report that includes the design YAML and the run log can be
|
|
195
|
+
acted on, because together they record the inputs and every step that ran.
|
|
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
|
|
|
4
4
|
|
|
5
5
|
[project]
|
|
6
6
|
name = "lexsync"
|
|
7
|
-
version = "0.
|
|
7
|
+
version = "0.2.0"
|
|
8
8
|
description = "Lexical optimisation and hardware-timed experiment generation"
|
|
9
9
|
readme = "README.md"
|
|
10
10
|
requires-python = ">=3.10"
|
|
@@ -20,13 +20,12 @@ keywords = [
|
|
|
20
20
|
"counterbalancing", "pseudoword generation", "EEG", "ERP",
|
|
21
21
|
"PsychoPy", "OpenSesame", "jsPsych", "FAIR",
|
|
22
22
|
]
|
|
23
|
-
# No "License :: ..." classifier belongs in this list
|
|
24
|
-
#
|
|
25
|
-
#
|
|
26
|
-
#
|
|
27
|
-
# file to copy this list from.
|
|
23
|
+
# No "License :: ..." classifier belongs in this list. PEP 639 deprecates licence
|
|
24
|
+
# classifiers and lets build backends reject them alongside a License-Expression
|
|
25
|
+
# (setuptools does). PyPI accepts the pairing today, but the expression above
|
|
26
|
+
# already says all a classifier would. tests/test_metadata.py keeps one out.
|
|
28
27
|
#
|
|
29
|
-
# The
|
|
28
|
+
# The six interpreter lines are the six the test matrix runs, on Linux, macOS
|
|
30
29
|
# and Windows alike, so each one states something CI has checked. Alpha is the
|
|
31
30
|
# trove equivalent of the experimental lifecycle badge the README carries, and it
|
|
32
31
|
# is what all four siblings declare. "Typing :: Typed" goes with the py.typed
|
|
@@ -42,6 +41,7 @@ classifiers = [
|
|
|
42
41
|
"Programming Language :: Python :: 3.12",
|
|
43
42
|
"Programming Language :: Python :: 3.13",
|
|
44
43
|
"Programming Language :: Python :: 3.14",
|
|
44
|
+
"Programming Language :: Python :: 3.15",
|
|
45
45
|
"Programming Language :: Python :: 3 :: Only",
|
|
46
46
|
"Topic :: Scientific/Engineering",
|
|
47
47
|
"Topic :: Scientific/Engineering :: Information Analysis",
|
|
@@ -80,8 +80,10 @@ corpora = ["wordfreq>=3.0,<4"] # frozen 3.x line: a stable ~2021 language snap
|
|
|
80
80
|
experiment = ["psychopy>=2024.1", "pyserial>=3.5"]
|
|
81
81
|
# streamlit backs test_apps.py: without it those tests importorskip themselves and
|
|
82
82
|
# the app they cover goes unguarded in CI, so it belongs with the rest of what the
|
|
83
|
-
# suite needs to run.
|
|
84
|
-
|
|
83
|
+
# suite needs to run. 1.49 is the first release whose st.dataframe and
|
|
84
|
+
# st.data_editor accept width="stretch", which the app passes in place of the
|
|
85
|
+
# deprecated use_container_width.
|
|
86
|
+
dev = ["pytest>=7.0", "build", "twine", "streamlit>=1.49"]
|
|
85
87
|
|
|
86
88
|
[project.scripts]
|
|
87
89
|
lexsync = "lexsync.cli:main"
|
|
@@ -113,8 +115,8 @@ target-version = "py310"
|
|
|
113
115
|
# scripts, which the host runs with Canvas, var, clock and send_trigger already
|
|
114
116
|
# in the namespace, and PsychoPy templates whose ALL_CAPS placeholders are
|
|
115
117
|
# substituted before anything executes. Linting them as importable Python
|
|
116
|
-
# reports every one of those as an undefined name
|
|
117
|
-
#
|
|
118
|
+
# reports every one of those as an undefined name, which is every F finding
|
|
119
|
+
# this package has, and each is a false positive by construction.
|
|
118
120
|
# They are held to their own standard instead: tests/test_templates.py and the R
|
|
119
121
|
# twin pin all three copies byte-identical to templates/ at the repository root.
|
|
120
122
|
extend-exclude = ["src/lexsync/templates"]
|
|
@@ -126,21 +128,19 @@ extend-exclude = ["src/lexsync/templates"]
|
|
|
126
128
|
#
|
|
127
129
|
# F, E9 and B: the rules that catch bugs rather than style. Undefined names,
|
|
128
130
|
# unused imports and variables, redefined functions, f-strings missing their
|
|
129
|
-
# placeholders, syntax errors, and the bugbear set.
|
|
130
|
-
#
|
|
131
|
-
# only go red for a real regression.
|
|
131
|
+
# placeholders, syntax errors, and the bugbear set. The package passes them, so
|
|
132
|
+
# a failure on them marks a real regression.
|
|
132
133
|
#
|
|
133
|
-
# B905
|
|
134
|
+
# B905 is the main reason B is enabled. Every zip() in this package pairs
|
|
134
135
|
# sequences that are equal in length by construction -- columns of one frame,
|
|
135
|
-
# or a list built from the thing it is zipped against -- so each
|
|
136
|
+
# or a list built from the thing it is zipped against -- so each says
|
|
136
137
|
# strict=True. That is a no-op while the invariant holds and an exception the
|
|
137
138
|
# moment it stops holding, which matters most in counterbalancing.py and
|
|
138
139
|
# scripting.py, where the zipped values build the hash keys behind the
|
|
139
140
|
# deterministic shuffle and the jitter. Silent truncation there would not raise;
|
|
140
|
-
# it would quietly deal a different experiment.
|
|
141
|
-
# the suite and a full pipeline run over 242 artefacts are byte-identical.
|
|
141
|
+
# it would quietly deal a different experiment.
|
|
142
142
|
#
|
|
143
|
-
# E, W, I and UP are enforced too, so
|
|
143
|
+
# E, W, I and UP are enforced too, so the package answers to the same rule set
|
|
144
144
|
# as its four siblings. Three rules are turned off, each because the thing it
|
|
145
145
|
# objects to is carrying meaning rather than costing anything:
|
|
146
146
|
#
|
|
@@ -173,12 +173,13 @@ packages = ["src/lexsync"]
|
|
|
173
173
|
#
|
|
174
174
|
# tests is here because the sdist is what conda-forge, Debian and Nix build from,
|
|
175
175
|
# and a packager who receives no suite can do nothing more than import the module
|
|
176
|
-
# and hope. The
|
|
176
|
+
# and hope. The files it adds include the tests that pin values the R twin
|
|
177
177
|
# computes, many of them bit for bit, and those are the only check a build from
|
|
178
178
|
# the sdist gets that this engine still agrees with R. That agreement is
|
|
179
179
|
# precisely what a rebuild on an unfamiliar platform might disturb. The
|
|
180
180
|
# comparisons of whole artefacts with the R engine's output, test_parity.py and
|
|
181
|
-
# test_byte_parity.py,
|
|
181
|
+
# test_byte_parity.py, and the twin-edge cases of test_twin_edge.py, need the
|
|
182
|
+
# rest of the repository and skip there, as does
|
|
182
183
|
# every other test that reads it. tests/helper_repo.py says how they tell the
|
|
183
184
|
# repository from whatever else surrounds the unpacked sdist.
|
|
184
185
|
#
|
|
@@ -186,7 +187,8 @@ packages = ["src/lexsync"]
|
|
|
186
187
|
# everything else that is tracked. That is shorter to write but wrong here: this
|
|
187
188
|
# directory also holds docs/, overrides/, mkdocs.yml, run_pipeline.py and three
|
|
188
189
|
# requirements files, none of which belong in a distribution, and every one of
|
|
189
|
-
# them would have to be listed to keep it out. An include list
|
|
190
|
-
#
|
|
191
|
-
#
|
|
190
|
+
# them would have to be listed to keep it out. An include list ships what it
|
|
191
|
+
# names, plus the repository's .gitignore, which hatchling always adds to an
|
|
192
|
+
# sdist, so forgetting an entry costs a missing file, and forgetting one in an
|
|
193
|
+
# exclude list costs a tarball quietly twice the size it should be.
|
|
192
194
|
include = ["src/lexsync", "tests", "README.md", "LICENSE", "LICENSE.note", "pyproject.toml"]
|
|
@@ -1,7 +1,8 @@
|
|
|
1
1
|
"""lexsync: lexical optimisation and hardware-timed experiment generation."""
|
|
2
|
-
from .corpora import fetch_corpus, list_corpora
|
|
2
|
+
from .corpora import fetch_corpus, list_corpora, normalise_lexicon
|
|
3
3
|
from .counterbalancing import balance_lists, counterbalance, participant_table
|
|
4
4
|
from .datasheet import build_datasheet, methods_paragraph, write_datasheet
|
|
5
|
+
from .design import design_stem, validate_design
|
|
5
6
|
from .generation import build_lexdec_stimuli, generate_pseudowords, make_pseudoword
|
|
6
7
|
from .matching import match_stimuli, resample_stimuli, select_continuous_stimuli
|
|
7
8
|
from .paradigms import PARADIGMS, required_fields, resolve_events
|
|
@@ -46,8 +47,8 @@ __all__ = [
|
|
|
46
47
|
"match_report_continuous", "export_experiments",
|
|
47
48
|
"export_psychopy", "export_opensesame", "export_jspsych", "resolve_trial_timing",
|
|
48
49
|
"add_pair_overlap",
|
|
49
|
-
"list_corpora", "fetch_corpus",
|
|
50
|
+
"list_corpora", "fetch_corpus", "normalise_lexicon",
|
|
50
51
|
"build_datasheet", "methods_paragraph", "write_datasheet",
|
|
51
|
-
"run_pipeline", "run_all",
|
|
52
|
+
"run_pipeline", "run_all", "validate_design", "design_stem",
|
|
52
53
|
]
|
|
53
|
-
__version__ = "0.
|
|
54
|
+
__version__ = "0.2.0"
|