def plot_places(min_prop=MIN_PROP):
place_counts = collections.Counter([
art['place']
for art in mongo['article'].find(
{},
{'place': True}
)
if art['place']
])
top_places = [
place[0]
for place in place_counts.most_common(10)
]
data = crosstab(
mongo['article'],
'place',
top_places,
'tags.label',
pattern.tag_groups['pkg'].labels
)
frame = pd.DataFrame(
data,
index=top_places,
columns=pattern.tag_groups['pkg'].labels,
)
min_count = min_prop * frame.sum().sum()
sum0 = frame.sum()
plot_frame = frame.ix[:, sum0 >= min_count]
plot_frame['other'] = frame.ix[:, sum0 < min_count].sum(axis=1)
plot_frame = plot_frame.div(
1.0 * plot_frame.sum(axis=1),
axis=0,
)
# Make plots
plot_tags_by_place(
plot_frame, stacked=True,
outname=file_name(['pkg', 'stacked'], path='place')
)
plot_tags_by_place(
plot_frame, stacked=False,
outname=file_name(['pkg', 'adjacent'], path='place')
)
def plot_places(min_prop=MIN_PROP):
place_counts = collections.Counter([
art['place'] for art in mongo['article'].find({}, {'place': True})
if art['place']
])
top_places = [place[0] for place in place_counts.most_common(10)]
data = crosstab(mongo['article'], 'place', top_places, 'tags.label',
pattern.tag_groups['pkg'].labels)
frame = pd.DataFrame(
data,
index=top_places,
columns=pattern.tag_groups['pkg'].labels,
)
min_count = min_prop * frame.sum().sum()
sum0 = frame.sum()
plot_frame = frame.ix[:, sum0 >= min_count]
plot_frame['other'] = frame.ix[:, sum0 < min_count].sum(axis=1)
plot_frame = plot_frame.div(
1.0 * plot_frame.sum(axis=1),
axis=0,
)
# Make plots
plot_tags_by_place(plot_frame,
stacked=True,
outname=file_name(['pkg', 'stacked'], path='place'))
plot_tags_by_place(plot_frame,
stacked=False,
outname=file_name(['pkg', 'adjacent'], path='place'))
def test_defaults(self):
assert_equal(utils.file_name(['foo', 'bar']),
os.path.join(trendpath.fig_dir, 'foo-bar'))
])
return dprimes, groups
print('{0} articles included in both sets'.format(
len(val.pmids)
))
print('{0} articles included in both sets, excluding supplements'.format(
len(val.pmids_no_supplement)
))
# Validate boolean values
validation = val.validate()
dprimes, groups = to_hist(validation)
val.validate_hist(
dprimes, labels=groups, title='Supplements included', xlabel='D-Prime',
outname=file_name(['dprime-supplements'], path='validate')
)
print('Categorical Validation: Supplements Included')
print(np.mean(sum(dprimes, [])))
validation = val.validate(no_supplement=True)
dprimes, groups = to_hist(validation)
val.validate_hist(
dprimes, labels=groups, title='Supplements excluded', xlabel='D-Prime',
outname=file_name(['dprime-no-supplements'], path='validate')
)
print('Categorical Validation: Supplements Excluded')
print(np.mean(sum(dprimes, [])))
# Validate continuous values
print('{0} articles included in both sets'.format(
len(val.pmids)
))
print('{0} articles included in both sets, excluding supplements'.format(
len(val.pmids_no_supplement)
))
# Validate categorical values
validation = val.validate()
dprimes, groups = to_hist(validation)
val.validate_hist(
dprimes, bins=10, labels=groups, title='Supplements included', xlabel='D-Prime',
outname=file_name(['dprime-supplements'], path='validate')
)
print('Categorical Validation: Supplements Included')
print(np.mean(sum(dprimes, [])))
validation = val.validate(no_supplement=True)
dprimes, groups = to_hist(validation)
import pdb; pdb.set_trace()
val.validate_hist(
dprimes, bins=10, labels=groups, title='Supplements excluded', xlabel='D-Prime',
outname=file_name(['dprime-no-supplements'], path='validate')
)
print('Categorical Validation: Supplements Excluded')
print(np.mean(sum(dprimes, [])))
summary, labels = summarize_ranks(
mongo['article'],
labels,
min_prop=0.05,
)
ranks = order.analyze_rank_order(summary, mongo['article'])
ranks_spm = order.analyze_rank_order(summary, mongo['article'],
{'tags.label': 'spm'})
ranks_fsl = order.analyze_rank_order(summary, mongo['article'],
{'tags.label': 'fsl'})
ranks_afni = order.analyze_rank_order(summary, mongo['article'],
{'tags.label': 'afni'})
seqplot.rank_plot(ranks, outname=file_name(['seq'], path='seq'))
seqplot.multi_rank_plot(ranks, [
'spm',
'fsl',
'afni',
], [
ranks_spm,
ranks_fsl,
ranks_afni,
],
outname=file_name(['seq', 'pkg'], path='seq'))
#years = range(2000, 2014)
#ranks_year = []
#
x_interp, interpolate.splev(x_interp, fit), '-',
)
ax = plt.gca()
ax.set_xlim((dates[0] - 1, dates[-1] + 1))
if title:
ax.set_title(title)
if xlabel:
ax.set_xlabel(xlabel)
if ylabel:
ax.set_ylabel(ylabel)
if outname:
plt.savefig(outname + '.pdf', bbox_inches='tight')
plt.close()
for group in pipeline.variance_groups:
plot_pipelines(
DATES, count_year[group], ylabel='Count',
outname=file_name(['count-{0}'.format(group)], path='pipelines')
)
plot_pipelines(
DATES, count_year[group], totals=total_year, ylabel='Proportion',
outname=file_name(['prop-{0}'.format(group)], path='pipelines')
)
print('{0} articles included in both sets'.format(len(val.pmids)))
print('{0} articles included in both sets, excluding supplements'.format(
len(val.pmids_no_supplement)))
# Validate categorical values
validation = val.validate()
dprimes, groups = to_hist(validation)
frame = to_frame(validation)
val.validate_hist(dprimes,
bins=5,
labels=groups,
title='Supplements included',
xlabel='D-Prime',
outname=file_name(['dprime-supplements'], path='validate'))
frame.to_csv(file_name(['dprime-supplements'], path='validate'), ext='.csv')
print('Categorical Validation: Supplements Included')
print(np.mean(sum(dprimes, [])))
validation = val.validate(no_supplement=True)
dprimes, groups = to_hist(validation)
frame = to_frame(validation)
val.validate_hist(dprimes,
bins=5,
labels=groups,
title='Supplements excluded',
xlabel='D-Prime',
outname=file_name(['dprime-no-supplements'],
path='validate'))
frame.to_csv(file_name(['dprime-no-supplements'], path='validate', ext='.csv'))
print('{0} articles included in both sets'.format(
len(val.pmids)
))
print('{0} articles included in both sets, excluding supplements'.format(
len(val.pmids_no_supplement)
))
# Validate categorical values
validation = val.validate()
dprimes, groups = to_hist(validation)
frame = to_frame(validation)
val.validate_hist(
dprimes, bins=5, labels=groups, title='Supplements included', xlabel='D-Prime',
outname=file_name(['dprime-supplements'], path='validate')
)
frame.to_csv(file_name(['dprime-supplements'], path='validate'), ext='.csv')
print('Categorical Validation: Supplements Included')
print(np.mean(sum(dprimes, [])))
validation = val.validate(no_supplement=True)
dprimes, groups = to_hist(validation)
frame = to_frame(validation)
val.validate_hist(
dprimes, bins=5, labels=groups, title='Supplements excluded', xlabel='D-Prime',
outname=file_name(['dprime-no-supplements'], path='validate')
)
frame.to_csv(file_name(['dprime-no-supplements'], path='validate', ext='.csv'))
print('Categorical Validation: Supplements Excluded')
def test_path(self):
assert_equal(
utils.file_name(['foo', 'bar'], path='bob'),
os.path.join(trendpath.fig_dir, 'bob', 'foo-bar')
)
def test_delim(self):
assert_equal(
utils.file_name(['foo', 'bar'], dlm='_'),
os.path.join(trendpath.fig_dir, 'foo_bar')
)
def test_defaults(self):
assert_equal(
utils.file_name(['foo', 'bar']),
os.path.join(trendpath.fig_dir, 'foo-bar')
)
"""
"""
from neurotrends.config import mongo
from neurotrends.model.utils import verified_mongo
from neurotrends.analysis.groupby.naive import summarize_custom
from neurotrends.analysis.groupby.naive import summarize_smooth_kernel, summarize_highpass_cutoff
from neurotrends.analysis.plot.histplot import hist
from neurotrends.analysis.plot.utils import file_name
import numpy as np
cursor = mongo['article'].find(verified_mongo, {'tags': 1, 'date': 1})
summary_smooth_kernel = summarize_custom(cursor, 'smooth_kernel',
summarize_smooth_kernel)
summary_highpass_cutoff = summarize_custom(cursor, 'highpass_cutoff',
summarize_highpass_cutoff)
hist(summary_smooth_kernel,
bins=np.arange(0.5, 19.5, 1),
xlabel='Smoothing Kernel',
outname=file_name(['smooth-kernel'], path='hist'))
hist(summary_highpass_cutoff,
xlog=True,
xlabel='High-pass Filter Cutoff',
outname=file_name(['highpass-cutoff'], path='hist'))
colors = get_colors(plot_frame.columns, PALETTE_NAME)
plot_frame = plot_frame.div(
1.0 * plot_frame.sum(axis=1),
axis=0,
)
plot_frame.plot(
kind='barh', stacked=True,
color=colors,
)
ax = plt.gca()
# Hack: Remove horizontal line at y=0 inserted by pandas
ax.lines.pop()
ax.invert_yaxis()
ax.set_xlabel('Proportion')
handles, labels = ax.get_legend_handles_labels()
lgd = ax.legend(
handles, labels,
loc='upper left', bbox_to_anchor=(1, 1)
)
plt.savefig(
file_name(['pkg', 'stacked'], path='place') + '.pdf',
bbox_inches='tight'
)
def test_delim(self):
assert_equal(utils.file_name(['foo', 'bar'], dlm='_'),
os.path.join(trendpath.fig_dir, 'foo_bar'))
def test_path(self):
assert_equal(utils.file_name(['foo', 'bar'], path='bob'),
os.path.join(trendpath.fig_dir, 'bob', 'foo-bar'))
"""
from neurotrends.config import mongo
from neurotrends.model.utils import verified_mongo
from neurotrends.analysis.groupby.naive import summarize_custom
from neurotrends.analysis.groupby.naive import summarize_smooth_kernel, summarize_highpass_cutoff
from neurotrends.analysis.plot.histplot import hist
from neurotrends.analysis.plot.utils import file_name
import numpy as np
cursor = mongo["article"].find(verified_mongo, {"tags": 1, "date": 1})
summary_smooth_kernel = summarize_custom(cursor, "smooth_kernel", summarize_smooth_kernel)
summary_highpass_cutoff = summarize_custom(cursor, "highpass_cutoff", summarize_highpass_cutoff)
hist(
summary_smooth_kernel,
bins=np.arange(0.5, 19.5, 1),
xlabel="Smoothing Kernel",
outname=file_name(["smooth-kernel"], path="hist"),
)
hist(
summary_highpass_cutoff,
xlog=True,
xlabel="High-pass Filter Cutoff",
outname=file_name(["highpass-cutoff"], path="hist"),
)
if label not in ['highpass_cutoff', 'smooth_kernel']
]
labels = list(set(labels))
summary, labels = summarize_ranks(
mongo['article'], labels, min_prop=0.05,
)
ranks = order.analyze_rank_order(summary, mongo['article'])
ranks_spm = order.analyze_rank_order(summary, mongo['article'], {'tags.label': 'spm'})
ranks_fsl = order.analyze_rank_order(summary, mongo['article'], {'tags.label': 'fsl'})
ranks_afni = order.analyze_rank_order(summary, mongo['article'], {'tags.label': 'afni'})
seqplot.rank_plot(ranks, outname=file_name(['seq'], path='seq'))
seqplot.multi_rank_plot(
ranks,
[
'spm',
'fsl',
'afni',
],
[
ranks_spm,
ranks_fsl,
ranks_afni,
],
outname=file_name(['seq', 'pkg'], path='seq')
)
ax.set_xlim((dates[0] - 1, dates[-1] + 1))
if title:
ax.set_title(title)
if xlabel:
ax.set_xlabel(xlabel)
if ylabel:
ax.set_ylabel(ylabel)
if outname:
plt.savefig(outname + '.pdf', bbox_inches='tight')
plt.close()
for group in pipeline.variance_groups:
plot_pipelines(DATES,
count_year[group],
ylabel='Count',
outname=file_name(['count-{0}'.format(group)],
path='pipelines'))
plot_pipelines(DATES,
count_year[group],
totals=total_year,
ylabel='Proportion',
outname=file_name(['prop-{0}'.format(group)],
path='pipelines'))