def evaluate_tempo_range(self,
seq_length=None,
query_duration=None,
tempo_range=[1.0],
gt_only=True):
"""
::
Loop over tempo range evaluating and integrating results:
query_duration = fractional seconds (requires adb.delta_time)
OR seq_length - integer length of query sequence
tempo_range = list of tempo proportions relative to nominal 1.0
gt_only = if True, return only ground-truth results otherwise return full database results
"""
seq_length = self.set_seq_length(seq_length, query_duration)
out_ranks = []
out_dists = []
for t in tempo_range:
print "Evaluating tempo: ", t
# perform full evaluation, gt_only=False, for across-tempo merged ranks of gt in db results
ranks, dists = self.evaluate(query_duration=query_duration,
tempo=t,
gt_only=False)
out_ranks.append(ranks)
out_dists.append(dists)
out_ranks = pylab.dstack(out_ranks).min(2) # collapse runs
out_dists = pylab.dstack(out_dists).min(2)
if gt_only:
out_ranks, out_dists = self.reduce_evaluation_to_gt(
out_ranks, out_dists, query_duration=query_duration)
return out_ranks, out_dists
def __update_graphics_context__(gc, scatter, pp_spec, manager, first=False):
"""
function update graphics context of scatter plot based in
information included in poincare plot specification object
"""
p = pp_spec # alias
m = manager # alias
if first and p.level >= 0 and p.inactive_stop >= 0 and p.active_stop >= 0:
r_points = pl.dstack(
(p.x_data[p.inactive_stop:p.active_stop],
p.y_data[p.inactive_stop:p.active_stop]))[0]
r_points = scatter.map_screen(r_points)
gc.set_fill_color(m.active_color_as_tuple)
gc.draw_marker_at_points(r_points, m.active_point_size, CIRCLE)
if p.inactive_start_2 >= 0 and p.inactive_stop_2 >= 0:
r_points = pl.dstack(
(p.x_data[p.inactive_start_2: p.inactive_stop_2],
p.y_data[p.inactive_start_2: p.inactive_stop_2]))[0]
r_points = scatter.map_screen(r_points)
gc.set_fill_color(m.inactive_color_as_tuple)
gc.draw_marker_at_points(r_points, m.inactive_point_size, CIRCLE)
if first and p.level >= 0 and p.inactive_stop >= 0:
r_points = pl.dstack(
(p.x_data[: p.inactive_stop],
p.y_data[: p.inactive_stop]))[0]
r_points = scatter.map_screen(r_points)
gc.set_fill_color(m.inactive_color_as_tuple)
gc.draw_marker_at_points(r_points, m.inactive_point_size, CIRCLE)
elif p.inactive_start >= 0 and p.inactive_stop >= 0:
r_points = pl.dstack(
(p.x_data[p.inactive_start: p.inactive_stop],
p.y_data[p.inactive_start: p.inactive_stop]))[0]
r_points = scatter.map_screen(r_points)
gc.set_fill_color(m.inactive_color_as_tuple)
gc.draw_marker_at_points(r_points, m.inactive_point_size, CIRCLE)
if (not first or p.level == 0) \
and p.active_start >= 0 and p.active_stop >= 0:
r_points = pl.dstack(
(p.x_data[p.active_start: p.active_stop],
p.y_data[p.active_start: p.active_stop]))[0]
r_points = scatter.map_screen(r_points)
gc.set_fill_color(m.active_color_as_tuple)
gc.draw_marker_at_points(r_points, m.active_point_size, CIRCLE)
def evaluate_tempo_range(self, seq_length=None, query_duration=None, tempo_range=[1.0], gt_only=True):
"""
::
Loop over tempo range evaluating and integrating results:
query_duration = fractional seconds (requires adb.delta_time)
OR seq_length - integer length of query sequence
tempo_range = list of tempo proportions relative to nominal 1.0
gt_only = if True, return only ground-truth results otherwise return full database results
"""
seq_length=self.set_seq_length(seq_length, query_duration)
out_ranks = []
out_dists = []
for t in tempo_range:
print "Evaluating tempo: ", t
# perform full evaluation, gt_only=False, for across-tempo merged ranks of gt in db results
ranks, dists = self.evaluate(query_duration=query_duration, tempo=t, gt_only=False)
out_ranks.append(ranks)
out_dists.append(dists)
out_ranks = pylab.dstack(out_ranks).min(2) # collapse runs
out_dists = pylab.dstack(out_dists).min(2)
if gt_only:
out_ranks, out_dists = self.reduce_evaluation_to_gt(out_ranks, out_dists, query_duration=query_duration)
return out_ranks, out_dists
