def test_fboxplot_rainbowplot(close_figures):
# Test fboxplot and rainbowplot together, is much faster.
def harmfunc(t):
"""Test function, combination of a few harmonic terms."""
# Constant, 0 with p=0.9, 1 with p=1 - for creating outliers
ci = int(np.random.random() > 0.9)
a1i = np.random.random() * 0.05
a2i = np.random.random() * 0.05
b1i = (0.15 - 0.1) * np.random.random() + 0.1
b2i = (0.15 - 0.1) * np.random.random() + 0.1
func = (1 - ci) * (a1i * np.sin(t) + a2i * np.cos(t)) + \
ci * (b1i * np.sin(t) + b2i * np.cos(t))
return func
np.random.seed(1234567)
# Some basic test data, Model 6 from Sun and Genton.
t = np.linspace(0, 2 * np.pi, 250)
data = [harmfunc(t) for _ in range(20)]
# fboxplot test
fig = plt.figure()
ax = fig.add_subplot(111)
_, depth, ix_depth, ix_outliers = fboxplot(data, wfactor=2, ax=ax)
ix_expected = np.array(
[13, 4, 15, 19, 8, 6, 3, 16, 9, 7, 1, 5, 2, 12, 17, 11, 14, 10, 0, 18])
assert_equal(ix_depth, ix_expected)
ix_expected2 = np.array([2, 11, 17, 18])
assert_equal(ix_outliers, ix_expected2)
# rainbowplot test (re-uses depth variable)
xdata = np.arange(data[0].size)
fig = rainbowplot(data, xdata=xdata, depth=depth, cmap=plt.cm.rainbow)
def test_fboxplot_rainbowplot(close_figures):
# Test fboxplot and rainbowplot together, is much faster.
def harmfunc(t):
"""Test function, combination of a few harmonic terms."""
# Constant, 0 with p=0.9, 1 with p=1 - for creating outliers
ci = int(np.random.random() > 0.9)
a1i = np.random.random() * 0.05
a2i = np.random.random() * 0.05
b1i = (0.15 - 0.1) * np.random.random() + 0.1
b2i = (0.15 - 0.1) * np.random.random() + 0.1
func = (1 - ci) * (a1i * np.sin(t) + a2i * np.cos(t)) + \
ci * (b1i * np.sin(t) + b2i * np.cos(t))
return func
np.random.seed(1234567)
# Some basic test data, Model 6 from Sun and Genton.
t = np.linspace(0, 2 * np.pi, 250)
data = [harmfunc(t) for _ in range(20)]
# fboxplot test
fig = plt.figure()
ax = fig.add_subplot(111)
_, depth, ix_depth, ix_outliers = fboxplot(data, wfactor=2, ax=ax)
ix_expected = np.array([13, 4, 15, 19, 8, 6, 3, 16, 9, 7, 1, 5, 2,
12, 17, 11, 14, 10, 0, 18])
assert_equal(ix_depth, ix_expected)
ix_expected2 = np.array([2, 11, 17, 18])
assert_equal(ix_outliers, ix_expected2)
# rainbowplot test (re-uses depth variable)
xdata = np.arange(data[0].size)
fig = rainbowplot(data, xdata=xdata, depth=depth, cmap=plt.cm.rainbow)
def main(argstr=None):
import argparse
# helper for boolean flags
# based on http://stackoverflow.com/a/9236426/344821
class ActionNoYes(argparse.Action):
def __init__(self,
opt_name,
off_name=None,
dest=None,
default=True,
required=False,
help=None):
if off_name is None:
off_name = 'no-' + opt_name
self.off_name = '--' + off_name
if dest is None:
dest = opt_name.replace('-', '_')
super(ActionNoYes,
self).__init__(['--' + opt_name, '--' + off_name],
dest,
nargs=0,
const=None,
default=default,
required=required,
help=help)
def __call__(self, parser, namespace, values, option_string=None):
setattr(namespace, self.dest, option_string != self.off_name)
parser = argparse.ArgumentParser()
parser.add_argument('files', nargs='+')
g = parser.add_mutually_exclusive_group()
g.add_argument('--over-random', action='store_true', default=False)
g.add_argument('--absolute', action='store_false', dest='over_random')
parser.add_argument('--key-regexes',
'--keys',
nargs='*',
metavar='RE',
default=[re.compile('.*')],
type=re.compile)
parser.add_argument('--key-exclude-regexes',
'--skip-keys',
nargs='*',
default=[],
type=re.compile,
metavar='RE')
parser.add_argument('--legend',
default='outside',
choices={'outside', 'inside'})
g = parser.add_argument_group('Plot Types')
g._add_action(ActionNoYes('rmses', default=False))
g._add_action(ActionNoYes('rmse-fboxplots', default=False))
g._add_action(ActionNoYes('auc', default=True))
g._add_action(ActionNoYes('predaucs', default=False))
g._add_action(ActionNoYes('predauc-auc', default=False))
g.add_argument('--ge-cutoff', nargs='+', type=float)
g.add_argument('--ge-cutoff-auc', nargs='+', type=float)
#parser.add_argument('--save')
if argstr is not None:
import shlex
args = parser.parse_args(shlex.split(argstr))
else:
args = parser.parse_args()
#if args.save:
# import matplotlib
# matplotlib.use('Agg')
import matplotlib.pyplot as plt
res = load_data(args.files,
do_rmse=args.rmses,
do_rmse_auc=args.auc,
do_predauc=args.predaucs,
do_predauc_auc=args.predauc_auc,
do_cutoffs=args.ge_cutoff,
do_cutoff_aucs=args.ge_cutoff_auc,
rmse_over_random=args.over_random,
ret_rmse_traces=args.rmse_fboxplots)
if args.rmse_fboxplots:
data, rmse_traces = res
else:
data = res
ns = data['ns']
#key_res = [re.compile(r) for r in args.key_regexes]
#key_bads = [re.compile(r) for r in args.key_exclude_regexes]
def filter_keys(d):
return {
k: v
for k, v in d.items()
if any(r.search(k) for r in args.key_regexes) and not any(
r.search(k) for r in args.key_exclude_regexes)
}
over_random = ' over_random' if args.over_random else ''
# rmse stuff
if args.rmses:
plt.figure()
plot_lines(ns, filter_keys(data['rmse']), 'RMSE' + over_random)
show_legend(args.legend)
if args.rmse_fboxplots:
from statsmodels.graphics.functional import fboxplot
for name, trace in filter_keys(rmse_traces).items():
fboxplot(trace, xdata=ns)
plt.hlines(0, *plt.xlim(), color='k')
plt.title(KEY_NAMES.get(name, name))
plt.xlabel("# of rated elements")
plt.ylabel('RMSE' + over_random)
if args.auc:
plt.figure()
plot_aucs(filter_keys(data['rmse_auc']),
'AUC ({})'.format('RMSE' + over_random))
# prediction auc stuff
if args.predaucs:
plt.figure()
plot_lines(ns, filter_keys(data['predauc']),
'Prediction AUC' + over_random)
show_legend(args.legend)
if args.auc:
plt.figure()
plot_aucs(filter_keys(data['predauc_auc']),
'AUC ({})'.format('Prediction AUC' + over_random))
# # >= cutoff stuff
if args.ge_cutoff:
for cutoff in args.ge_cutoff:
plt.figure()
plot_lines(ns, filter_keys(data['cutoffs'][cutoff]),
'# >= {}'.format(cutoff))
show_legend(args.legend)
if args.ge_cutoff_auc:
for cutoff in args.ge_cutoff_auc:
plt.figure()
plot_aucs(filter_keys(data['cutoff_aucs'][cutoff]),
'AUC (# >= {})'.format(cutoff))
plt.show()