def generate_session_pool(language="EN"):
"""Generate a single session pool for catFR experiments.
:param str language: Language to load words in.
:param int num_lists: Number of lists to assign.
:param int listno_start: Where to start numbering from.
:returns: Shuffled, categorized word pool.
"""
# validate language
if language.lower() not in ["en", "sp"]:
raise exc.LanguageError(
"Language must be 'EN' or 'SP'".format(language))
# Load and shuffle order of words in categories
filename = "ram_categorized_{:s}.txt".format(language.lower())
pool = wordpool.shuffle_within_groups(
wordpool.load(filename),
"category")
words = sort_pairs(assign_list_numbers(assign_word_numbers(pool)))
all_words = wordpool.load('practice_cat_{:s}.txt'.format(language.lower()))
all_words['listno'] = 0
all_words['type'] = "PRACTICE"
all_words['wordno'] = range(12)
all_words['category'] = "X"
all_words['category_num'] = -999
all_words = wordpool.shuffle_words(all_words)
return all_words.append(words).reset_index(drop=True)
def test_create(pool, catpool):
assert "word" in pool
assert "word" in catpool
assert "category" in catpool
output = subprocess.check_output("git rev-parse --show-toplevel".split())
root = output.decode().strip()
other = wordpool.load(osp.join(root, "wordpool", "data", "ram_wordpool_en.txt"), False)
assert "word" in other
def test_write_wordpool(tmpdir):
fn = listgen.write_wordpool_txt
# unsupported language
with pytest.raises(exc.LanguageError):
fn(tmpdir, "DA")
# missing rec words for supported language
with pytest.raises(exc.LanguageError):
fn(tmpdir, "SP", True)
# writing without lures
with subdir(tmpdir) as path:
ret = fn(path, "EN")
assert len(os.listdir(path)) == 1
assert len(ret) == 1
assert ret[0] == osp.join(path, "RAM_wordpool.txt")
with open(ret[0]) as f:
words = pd.Series([l.strip() for l in f.readlines()])
assert (words == wordpool.load("ram_wordpool_en.txt").word).all()
# Writing with lures
with subdir(tmpdir) as path:
ret = fn(path, "EN", True)
assert len(os.listdir(path)) == 2
assert len(ret) == 2
assert ret[0] == osp.join(path, "RAM_wordpool.txt")
assert ret[1] == osp.join(path, "RAM_lurepool.txt")
# targets
with open(ret[0]) as f:
words = pd.Series([l.strip() for l in f.readlines()])
assert (words == wordpool.load("ram_wordpool_en.txt").word).all()
# lures
with open(ret[1]) as f:
words = pd.Series([l.strip() for l in f.readlines()])
assert (words == wordpool.load("REC1_lures_en.txt").word).all()
def test_generate_rec1_blocks(self):
pool = listgen.fr.generate_session_pool()
assigned = listgen.assign_list_types(pool, 3, 6, 16, 0)
lures = wordpool.load("REC1_lures_en.txt")
blocks = listgen.generate_rec1_blocks(assigned, lures)
assert isinstance(blocks, pd.DataFrame)
assert not all(
[s == u for s, u in zip(sorted(blocks.listno), blocks.listno)])
blocks2 = listgen.generate_rec1_blocks(assigned, lures)
for n in range(len(blocks.word)):
if blocks.word.iloc[0] != blocks2.word.iloc[0]:
break
assert n < len(blocks.word)
# this should be the original index before being reset
assert "index" in blocks.columns
def catpool(language="en"):
yield wordpool.load("ram_categorized_{:s}.txt".format(language))
def pool(language="en"):
yield wordpool.load("ram_wordpool_{:s}.txt".format(language))
def catpool(self):
return wordpool.load("ram_categorized_en.txt")
"""FR list generation."""
import pandas as pd
import wordpool
RAM_LIST_EN = wordpool.load("ram_wordpool_en.txt")
RAM_LIST_SP = wordpool.load("ram_wordpool_sp.txt")
PRACTICE_LIST_EN = wordpool.load("practice_en.txt")
PRACTICE_LIST_SP = wordpool.load("practice_sp.txt")
def generate_session_pool(words_per_list=12, num_lists=25, language="EN"):
"""Generate the pool of words for a single task session. This does *not*
assign stim, no-stim, or PS metadata since this part depends on the
experiment.
:param int words_per_list: Number of words in each list.
:param int num_lists: Total number of lists excluding the practice list.
:param str language: Session language (``EN`` or ``SP``).
:returns: Word pool
:rtype: pd.DataFrame
"""
assert language in ("EN", "SP")
practice = PRACTICE_LIST_EN if language == "EN" else PRACTICE_LIST_SP
practice["type"] = "PRACTICE"
practice["listno"] = 0
practice = wordpool.shuffle_words(practice).reset_index(drop=True)
"""List generation and I/O."""
import random
import os.path as osp
import numpy.random as npr
import pandas as pd
import wordpool
from .. import exc
from . import fr
from . import catfr
from . import pal
RAM_LIST_EN = wordpool.load("ram_wordpool_en.txt")
RAM_LIST_SP = wordpool.load("ram_wordpool_sp.txt")
CAT_LIST_EN = wordpool.load("ram_categorized_en.txt")
CAT_LIST_SP = wordpool.load("ram_categorized_sp.txt")
PRACTICE_LIST_EN = wordpool.load("practice_en.txt")
PRACTICE_LIST_SP = wordpool.load("practice_sp.txt")
LURES_LIST_EN = wordpool.load("REC1_lures_en.txt")
def write_wordpool_txt(path,
language="EN",
include_lure_words=False,
categorized=False):
"""Write `RAM_wordpool.txt` or `CatFR_WORDS.txt` to a file (why the naming
is so inconsistent is beyond me). This is used in event post-processing.