def vizualize_all_threshold_experiments():
thresh_output_dir = join_mkdir(results_root, 'threshold_matches'+hs_configstr)
exptx = 0
for exptx in range(len(vsdb_list)):
# Database handles.
hsA, hsB = vsdb_list[exptx]
expt_lbl = [within_lbl, cross_lbl][not hsA is hsB]
results_name = get_results_name(hsA, hsB)
print(' --- ---')
print(' - database ='+results_name+' ---')
print(' - expt_lbl ='+expt_lbl+' ---')
count2rr_AB = count2rr_list[exptx]
if hsA is hsB or not sansgt:
# Visualize chips which have a results with a high score
thresh_out_dir = join_mkdir(thresh_output_dir, expt_lbl)
viz_threshold_matchings(hsA, hsB, count2rr_AB, thresh_out_dir)
def vizualize_all_threshold_experiments():
thresh_output_dir = join_mkdir(results_root,
'threshold_matches' + hs_configstr)
exptx = 0
for exptx in range(len(vsdb_list)):
# Database handles.
hsA, hsB = vsdb_list[exptx]
expt_lbl = [within_lbl, cross_lbl][not hsA is hsB]
results_name = get_results_name(hsA, hsB)
print(' --- ---')
print(' - database =' + results_name + ' ---')
print(' - expt_lbl =' + expt_lbl + ' ---')
count2rr_AB = count2rr_list[exptx]
if hsA is hsB or not sansgt:
# Visualize chips which have a results with a high score
thresh_out_dir = join_mkdir(thresh_output_dir, expt_lbl)
viz_threshold_matchings(hsA, hsB, count2rr_AB, thresh_out_dir)
def viz_threshold_matchings(hsA, hsB, count2rr_AB, thresh_out_dir):
'returns database, cx, database cx'
import numpy as np
valid_cxsB = hsB.cm.get_valid_cxs()
num_matching = 0
MATCH_THRESHOLD = __method_2_matchthresh__.get(__METHOD__, 10)
results_name = get_results_name(hsA, hsB)
print(' * Visualizing threshold matchings ' + results_name +
' give it some time to plot...')
threshdb_out_dir = join_mkdir(thresh_out_dir, results_name)
# For each query run in hsA vs hsB
for count in xrange(len(count2rr_AB)):
rr = count2rr_AB[count]
qcx = rr.qcx
res = QueryResult(hsB, rr, hsA)
qname = res.qhs.cm.cx2_name(qcx)
# Set matching threshold
res.top_thresh = MATCH_THRESHOLD
res.num_top_min = 0
res.num_top_max = 5
res.num_extra_return = 0
# Check to see if any matched over the threshold
top_cxs = res.top_cx()
top_names = res.hs.cm.cx2_name(top_cxs)
top_scores = res.scores()[top_cxs]
if len(top_cxs) > 0:
# Visualize the result
# Stupid segfaults #if qcx == 113: #41: #import IPython #IPython.embed() # Get the scores
num_matching += 1
res.visualize()
# Create a filename showing dataset, score, cx, match names
matchname_set = \
set([name.replace('Lionfish','') for name in (top_names+[qname])])
matchnames = '-'.join(list(matchname_set))
scorestr = str(int(round(top_scores[0])))
fig_fname = '-'.join([
results_name, 'score' + scorestr, 'cx' + str(qcx),
'MATCHES' + matchnames
]) + '.jpg'
fig = myfigure(0)
fig_fpath = join(threshdb_out_dir, fig_fname)
sys.stdout.write('.')
safe_savefig(fig, fig_fpath)
print(' * Visualized %d above thresh: %f from expt: %s ' %
(num_matching, MATCH_THRESHOLD, results_name))
def viz_threshold_matchings(hsA, hsB, count2rr_AB, thresh_out_dir):
'returns database, cx, database cx'
import numpy as np
valid_cxsB = hsB.cm.get_valid_cxs()
num_matching = 0
MATCH_THRESHOLD = __method_2_matchthresh__.get(__METHOD__, 10)
results_name = get_results_name(hsA, hsB)
print(' * Visualizing threshold matchings '+results_name+' give it some time to plot...')
threshdb_out_dir = join_mkdir(thresh_out_dir, results_name)
# For each query run in hsA vs hsB
for count in xrange(len(count2rr_AB)):
rr = count2rr_AB[count]
qcx = rr.qcx
res = QueryResult(hsB, rr, hsA)
qname = res.qhs.cm.cx2_name(qcx)
# Set matching threshold
res.top_thresh = MATCH_THRESHOLD
res.num_top_min = 0
res.num_top_max = 5
res.num_extra_return = 0
# Check to see if any matched over the threshold
top_cxs = res.top_cx()
top_names = res.hs.cm.cx2_name(top_cxs)
top_scores = res.scores()[top_cxs]
if len(top_cxs) > 0:
# Visualize the result
# Stupid segfaults #if qcx == 113: #41: #import IPython #IPython.embed() # Get the scores
num_matching += 1
res.visualize()
# Create a filename showing dataset, score, cx, match names
matchname_set = \
set([name.replace('Lionfish','') for name in (top_names+[qname])])
matchnames = '-'.join(list(matchname_set))
scorestr = str(int(round(top_scores[0])))
fig_fname = '-'.join([results_name,
'score'+scorestr,
'cx'+str(qcx),
'MATCHES'+matchnames])+'.jpg'
fig = myfigure(0)
fig_fpath = join(threshdb_out_dir, fig_fname)
sys.stdout.write('.')
safe_savefig(fig, fig_fpath)
print(' * Visualized %d above thresh: %f from expt: %s ' % (num_matching,
MATCH_THRESHOLD,
results_name))
def vizualize_all_chipscores():
chipscore_dir = join_mkdir(results_root,
'chipscore_frequencies' + hs_configstr)
if sansgt:
Rank1TNFig = myfigure(
200,
doclf=True,
title='Frequency of true negative chip scores')
CrossAndTNFig = myfigure(
400,
doclf=True,
title='True negatives and cross database queries')
CrossFig = myfigure(
300,
doclf=True,
title='Frequency of all cross-database chip scores')
elif not sansgt:
AllTPFig = myfigure(100,
doclf=True,
title='Frequency of true positive chip scores')
cross_db_scores = []
true_pos_scores = []
cmap = plt.get_cmap('Set1')
for exptx in range(total_expts):
# Database handles.
hsA, hsB = vsdb_list[exptx]
expt_lbl = [within_lbl, cross_lbl][not hsA is hsB]
expt_color = cmap(exptx / float(total_expts))
results_name = get_results_name(hsA, hsB)
results_name_abrev = get_results_name(hsA, hsB, abrev=True)
print(' --- ---')
print(' - database =' + results_name + ' ---')
print(' - expt_lbl =' + expt_lbl + ' ---')
count2rr_AB = count2rr_list[exptx]
print(' * Visualizing chip score frequencies ' + results_name)
chipscore_data, ischipscore_TP = get_chipscores(
hsA, hsB, count2rr_AB)
txt_output_fpath = sanatize_fpath(join(results_root,
results_name_abrev+\
'_allscores_'+\
results_configstr+'.txt'))
write_all_chipscore_results(hsA, hsB, count2rr_AB,
txt_output_fpath)
if sansgt and hsA is hsB:
print(' * rank1 not TP')
# First true negative - within db (sansgt)
viz_chipscores(chipscore_data,
chipscore_mask=True - ischipscore_TP,
fig=Rank1TNFig,
holdon=True,
color=expt_color,
labelaug=results_name,
conditions='Rank=1, not TP')
viz_chipscores(chipscore_data,
chipscore_mask=True - ischipscore_TP,
fig=CrossAndTNFig,
holdon=True,
color=expt_color,
labelaug=results_name,
conditions='Rank=1, not TP')
elif sansgt and not hsA is hsB:
print(' * Rank <=__RESTRICT_TP__')
top_scores = chipscore_data[:, 0]
cross_db_scores.append(top_scores)
viz_chipscores(chipscore_data,
fig=CrossAndTNFig,
color=expt_color,
holdon=True,
labelaug=results_name,
conditions='Rank<=%d' % __RESTRICT_TP__)
if not sansgt and hsA is hsB:
# ALL TRUE POSITIVES - within a database
print(' * Top __RESTRICT_TP__ true positive chipscores')
top_scores = chipscore_data[:, 0:__RESTRICT_TP__]
top_mask = ischipscore_TP[:, 0:__RESTRICT_TP__]
true_pos_scores.append(top_scores[top_mask])
viz_chipscores(chipscore_data,
chipscore_mask=ischipscore_TP,
fig=AllTPFig,
holdon=True,
color=expt_color,
labelaug=results_name,
conditions='Rank<=%d, TP' % __RESTRICT_TP__)
if sansgt and not hsA is hsB:
# Highest Interdatabase matches for each combination of db
print(' * Top __RESTRICT_TP__ cross database chipscores')
top_scores = chipscore_data[:, 0]
cross_db_scores.append(top_scores)
viz_chipscores(chipscore_data,
fig=CrossFig,
color=expt_color,
holdon=True,
labelaug=results_name,
conditions='Rank<=%d' % __RESTRICT_TP__)
if sansgt:
print(' * Saving sansgt')
figfpath = join(chipscore_dir, within_lbl + '-rank1-chipscore')
safe_savefig(Rank1TNFig, figfpath, adjust_axes=True, trunc_max=100)
figfpath = join(chipscore_dir, within_lbl + '-and-cross-chipscore')
safe_savefig(CrossAndTNFig,
figfpath,
adjust_axes=True,
trunc_max=100)
safe_savefig(CrossFig,
join(chipscore_dir, 'crossdb-all-chipscores'),
adjust_axes=True,
trunc_max=100)
elif not sansgt:
print(' * Saving gt')
if len(cross_db_scores) > 0:
highest_cdscores = np.sort(
np.hstack(cross_db_scores))[::-1][0:20]
for c in highest_cdscores:
print('There are %d/%d TPs with scores less than %d' %
(np.sum(np.hstack(true_pos_scores) < c),
np.hstack(true_pos_scores).size, c))
# Finalize Plots and save
safe_savefig(
AllTPFig,
join(chipscore_dir,
within_lbl + '-top%dtp-chipscore' % __RESTRICT_TP__),
adjust_axes=True,
trunc_max=1000)
if dbtup[0] is dbtup[1]:
print(' %d --- sx%d - %s vs self' %
(dbx, symx, dbtup[0].get_dbid()))
else:
print(' %d --- sx%d - %s vs %s' %
(dbx, symx, dbtup[0].get_dbid(), dbtup[1].get_dbid()))
print_vsdb(vsdb_list, sym_list)
if __SANSGT__:
print(
'--- Database versus list __SANSGT__ ---\n DBX --- SYMX - hsA vs hsB '
)
print_vsdb(vsdb_sansgt_list, sym_sansgt_list)
# Dependents of parameters
results_root = join_mkdir('Results')
if not __cmd_mode__:
# Compute / Load all query results. Then visualize
print('Viz GT')
visualize_all_results(vsdb_list,
count2rr_list,
sym_list,
results_root,
sansgt=False)
if __SANSGT__:
print('Viz No GT')
visualize_all_results(vsdb_sansgt_list,
count2rr_list_sansgt,
sym_sansgt_list,
results_root,
def vizualize_all_chipscores():
chipscore_dir = join_mkdir(results_root, 'chipscore_frequencies'+hs_configstr)
if sansgt:
Rank1TNFig = myfigure(200, doclf=True,
title='Frequency of true negative chip scores')
CrossAndTNFig = myfigure(400, doclf=True,
title='True negatives and cross database queries')
CrossFig = myfigure(300, doclf=True,
title='Frequency of all cross-database chip scores')
elif not sansgt:
AllTPFig = myfigure(100, doclf=True,
title='Frequency of true positive chip scores')
cross_db_scores = []
true_pos_scores = []
cmap = plt.get_cmap('Set1')
for exptx in range(total_expts):
# Database handles.
hsA, hsB = vsdb_list[exptx]
expt_lbl = [within_lbl, cross_lbl][not hsA is hsB]
expt_color = cmap(exptx/float(total_expts))
results_name = get_results_name(hsA, hsB)
results_name_abrev = get_results_name(hsA, hsB, abrev=True)
print(' --- ---')
print(' - database ='+results_name+' ---')
print(' - expt_lbl ='+expt_lbl+' ---')
count2rr_AB = count2rr_list[exptx]
print(' * Visualizing chip score frequencies '+results_name)
chipscore_data, ischipscore_TP = get_chipscores(hsA, hsB, count2rr_AB)
txt_output_fpath = sanatize_fpath(join(results_root,
results_name_abrev+\
'_allscores_'+\
results_configstr+'.txt'))
write_all_chipscore_results(hsA, hsB, count2rr_AB, txt_output_fpath)
if sansgt and hsA is hsB:
print(' * rank1 not TP')
# First true negative - within db (sansgt)
viz_chipscores(chipscore_data, chipscore_mask=True - ischipscore_TP,
fig=Rank1TNFig, holdon=True,
color=expt_color, labelaug=results_name,
conditions='Rank=1, not TP')
viz_chipscores(chipscore_data, chipscore_mask=True - ischipscore_TP,
fig=CrossAndTNFig, holdon=True,
color=expt_color, labelaug=results_name,
conditions='Rank=1, not TP')
elif sansgt and not hsA is hsB:
print(' * Rank <=__RESTRICT_TP__')
top_scores = chipscore_data[:,0]
cross_db_scores.append(top_scores)
viz_chipscores(chipscore_data, fig=CrossAndTNFig,
color=expt_color, holdon=True,
labelaug=results_name, conditions='Rank<=%d' % __RESTRICT_TP__)
if not sansgt and hsA is hsB:
# ALL TRUE POSITIVES - within a database
print(' * Top __RESTRICT_TP__ true positive chipscores')
top_scores = chipscore_data[:,0:__RESTRICT_TP__]
top_mask = ischipscore_TP[:,0:__RESTRICT_TP__]
true_pos_scores.append(top_scores[top_mask])
viz_chipscores(chipscore_data, chipscore_mask=ischipscore_TP,
fig=AllTPFig, holdon=True,
color=expt_color, labelaug=results_name,
conditions='Rank<=%d, TP' % __RESTRICT_TP__)
if sansgt and not hsA is hsB:
# Highest Interdatabase matches for each combination of db
print(' * Top __RESTRICT_TP__ cross database chipscores')
top_scores = chipscore_data[:,0]
cross_db_scores.append(top_scores)
viz_chipscores(chipscore_data, fig=CrossFig,
color=expt_color, holdon=True,
labelaug=results_name, conditions='Rank<=%d' % __RESTRICT_TP__)
if sansgt:
print(' * Saving sansgt')
figfpath=join(chipscore_dir, within_lbl+'-rank1-chipscore')
safe_savefig(Rank1TNFig,figfpath, adjust_axes=True,trunc_max=100)
figfpath=join(chipscore_dir, within_lbl+'-and-cross-chipscore')
safe_savefig(CrossAndTNFig,figfpath,
adjust_axes=True,trunc_max=100)
safe_savefig(CrossFig,
join(chipscore_dir, 'crossdb-all-chipscores'),
adjust_axes=True, trunc_max=100)
elif not sansgt:
print(' * Saving gt')
if len(cross_db_scores) > 0:
highest_cdscores = np.sort(np.hstack(cross_db_scores))[::-1][0:20]
for c in highest_cdscores:
print('There are %d/%d TPs with scores less than %d' %
(np.sum(np.hstack(true_pos_scores) < c ),
np.hstack(true_pos_scores).size, c))
# Finalize Plots and save
safe_savefig(AllTPFig,
join(chipscore_dir, within_lbl+'-top%dtp-chipscore' % __RESTRICT_TP__),
adjust_axes=True, trunc_max=1000)
# List the database list we are running on
def print_vsdb(vsdb_list, sym_list):
print('--- Database versus list GT ---\n DBX --- SYMX - hsA vs hsB ')
for dbx, (dbtup, symx) in enumerate(zip(vsdb_list, sym_list)):
if dbtup[0] is dbtup[1]:
print(' %d --- sx%d - %s vs self' %
(dbx, symx, dbtup[0].get_dbid()) )
else:
print(' %d --- sx%d - %s vs %s' % (dbx, symx, dbtup[0].get_dbid(), dbtup[1].get_dbid()) )
print_vsdb(vsdb_list, sym_list)
if __SANSGT__:
print('--- Database versus list __SANSGT__ ---\n DBX --- SYMX - hsA vs hsB ')
print_vsdb(vsdb_sansgt_list, sym_sansgt_list)
# Dependents of parameters
results_root = join_mkdir('Results')
if not __cmd_mode__:
# Compute / Load all query results. Then visualize
print('Viz GT')
visualize_all_results(vsdb_list,
count2rr_list,
sym_list,
results_root,
sansgt=False)
if __SANSGT__:
print('Viz No GT')
visualize_all_results(vsdb_sansgt_list,
count2rr_list_sansgt,
sym_sansgt_list,
results_root,
