def test_dataset_iterator(self):
"""Dataset must correctly work as iterable object."""
dts = Dataset()
N = 10
for i in range(N):
dts.add_row([i])
for i in range(2):
self.assertEqual(N, len([i for i in dts]))
def test_dataset_iterator(self):
"""Dataset must correctly work as iterable object."""
dts = Dataset()
N = 10
for i in range(N):
dts.add_row([i])
for i in range(2):
self.assertEqual(N, len([i for i in dts]))
def test_load_invalid_date_type(self):
"""Test invalid data for dataset fails with verbose message."""
dts = Dataset([Dataset.DATE])
try:
dts.load([['aa-bb-cc']])
self.fail('dataset has loaded invalid data')
except dts.ParseError as exc:
self.assertEqual(
'Invalid value "aa-bb-cc" in line 1 '
'for "d" column type (index: 0)', str(exc))
def test_load_invalid_date_type(self):
"""Test invalid data for dataset fails with verbose message."""
dts = Dataset([Dataset.DATE])
try:
dts.load([['aa-bb-cc']])
self.fail('dataset has loaded invalid data')
except dts.ParseError as exc:
self.assertEqual('Invalid value "aa-bb-cc" in line 1 '
'for "d" column type (index: 0)',
str(exc))
def test_get_dataset_columns(self):
"""Test column splitting correctness.
| 0 | 1 | 2 |
| 3 | 4 | 5 |
"""
dts = Dataset()
[dts.add_row(i) for i in chunk(range(6), 3)]
for i in range(3):
self.assertEqual([i, i + 3], list(dts.column(i)))
# test multiple
self.assertEqual([(0, 1), (3, 4)], list(dts.column(0, 1)))
self.assertEqual([(1, 2), (4, 5)], list(dts.column(1, 2)))
def test_get_dataset_columns(self):
"""Test column splitting correctness.
| 0 | 1 | 2 |
| 3 | 4 | 5 |
"""
dts = Dataset()
[dts.add_row(i) for i in chunk(range(6), 3)]
for i in range(3):
self.assertEqual([i, i + 3], list(dts.column(i)))
# test multiple
self.assertEqual([(0, 1), (3, 4)], list(dts.column(0,1)))
self.assertEqual([(1, 2), (4, 5)], list(dts.column(1,2)))
def __init__(self, dataset):
self.mean = mean(dataset)
self.std = std(dataset, self.mean)
self.percentiles = percentiles(dataset, PERCENTILES)
self.calculated_percentiles = [
self.mean - 2 * self.std, self.mean - self.std,
self.mean - 0.67 * self.std, self.mean,
self.mean + 0.67 * self.std, self.mean + self.std,
self.mean + 2 * self.std
]
self.max = max(Dataset.get_num_column_or_list(dataset))
self.min = min(Dataset.get_num_column_or_list(dataset))
self.is_normal = self._is_normal(self.percentiles,
self.calculated_percentiles)
self.length = len(dataset)
def hist(dataset, **kwargs):
defaults = {
'grid': False,
'xlabel': '',
'ylabel': '',
'title': '',
'output': None,
}
graph_params = {
'bins': 20,
'normed': 1,
'facecolor': 'green',
'alpha': 0.75,
}
graph_params.update(kwargs)
defaults.update(kwargs)
[graph_params.pop(k) for k in defaults if k in graph_params]
values = list(Dataset.get_num_column_or_list(dataset))
n, bins, patches = plt.hist(values, **graph_params)
plt.xlabel(defaults['xlabel'])
plt.ylabel(defaults['ylabel'])
plt.title(defaults['title'])
plt.grid(defaults['grid'])
filename = defaults['output'] or get_tmp_file_name('.png')
plt.savefig(filename)
return filename
def hist(dataset, **kwargs):
defaults = {
'grid': False,
'xlabel': '',
'ylabel': '',
'title': '',
'output': None,
}
plot_params = {
'bins': 20,
'normed': 1,
'facecolor': 'green',
'alpha': 0.75,
}
plot_params.update(kwargs)
[plot_params.pop(k) for k in defaults if k in plot_params]
defaults.update(kwargs)
values = list(Dataset.get_num_column_or_list(dataset))
n, bins, patches = plt.hist(values, **plot_params)
plt.xlabel(defaults['xlabel'])
plt.ylabel(defaults['ylabel'])
plt.title(defaults['title'])
plt.grid(defaults['grid'])
filename = defaults['output'] or get_tmp_file_name('.png')
plt.savefig(filename)
return filename
def __init__(self, dataset):
self.mean = mean(dataset)
self.std = std(dataset, self.mean)
self.percentiles = percentiles(dataset, PERCENTILES)
self.calculated_percentiles = [
self.mean - 2 * self.std,
self.mean - self.std,
self.mean - 0.67 * self.std,
self.mean,
self.mean + 0.67 * self.std,
self.mean + self.std,
self.mean + 2 * self.std]
self.max = max(Dataset.get_num_column_or_list(dataset))
self.min = min(Dataset.get_num_column_or_list(dataset))
self.is_normal = self._is_normal(
self.percentiles, self.calculated_percentiles)
self.length = len(dataset)
def show_metric(metric):
conn = sqlite3.connect(DB)
points = get_points(conn, metric)
conn.close()
if not os.path.exists(IMAGE_DIR):
os.makedirs(IMAGE_DIR)
fname = IMAGE_DIR + '/%s.png' % metric
d = Dataset([Dataset.DATE, Dataset.FLOAT])
d.load(points)
plot_date(d, output=fname, figsize=(14, 7), linestyle='-')
data = open(fname, 'rb').read()
r = Response()
r.body = data
r.code = '200 OK'
r.headers = [('content-type', 'image/png'),
('content-length', str(len(r.body)))]
return r
def mean(dataset):
values = Dataset.get_num_column_or_list(dataset)
return sum(values) / float(len(dataset))
def std(dataset, m=None):
n = len(dataset)
values = Dataset.get_num_column_or_list(dataset)
if m is None:
m = mean(dataset)
return (sum((i - m)**2 for i in values) / float(n))**0.5
def mean(dataset):
values = Dataset.get_num_column_or_list(dataset)
return sum(values) / float(len(dataset))
def percentiles(dataset, vals):
n = len(dataset)
values = list(Dataset.get_num_column_or_list(dataset))
values.sort()
return [_percentile(values, n, p) for p in vals]
def std(dataset, m=None):
n = len(dataset)
values = Dataset.get_num_column_or_list(dataset)
if m is None:
m = mean(dataset)
return (sum((i - m) ** 2 for i in values) / float(n)) ** 0.5
def percentiles(dataset, vals):
n = len(dataset)
values = list(Dataset.get_num_column_or_list(dataset))
values.sort()
return [_percentile(values, n, p) for p in vals]