def dimension(self, expand=(0, 0, 0, 0), limits=None):
"""
The phyical size of the scale, if a position scale
Unlike limits, this always returns a numeric vector of length 2
"""
if limits is None:
limits = self.limits
c_range = self.range_c.range
d_range = self.limits
if self.is_empty():
return (0, 1)
elif self.range.range is None: # only continuous
return expand_range_distinct(c_range, expand)
elif c_range is None: # only discrete
# FIXME: I think this branch should not exist
return expand_range_distinct((1, len(d_range)), expand)
else: # both
# e.g categorical bar plot have discrete items, but
# are plot on a continuous x scale
a = np.hstack([
c_range,
expand_range_distinct((1, len(d_range)), expand)
])
return a.min(), a.max()
def _expand_range_distinct(x, expand):
# Expand ascending and descending order range
if x[0] > x[1]:
x = expand_range_distinct(x[::-1], expand)[::-1]
else:
x = expand_range_distinct(x, expand)
return x
def train(scale, limits, name):
"""
Train a single coordinate axis
"""
# Which axis are we dealing with
name = scale.aesthetics[0]
if self.expand:
expand = self.expand_default(scale)
else:
expand = (0, 0, 0, 0)
if limits is None:
rangee = scale.dimension(expand)
else:
rangee = scale.transform(limits)
rangee = expand_range_distinct(rangee, expand)
out = scale.break_info(rangee)
# This is where
# x_major, x_labels, x_minor, ...
# range keys are created
for key in list(out.keys()):
new_key = '{}_{}'.format(name, key)
out[new_key] = out.pop(key)
return out
def train(scale, limits, trans, name):
"""
Train a single coordinate axis
"""
if limits is None:
rangee = scale.dimension()
else:
rangee = scale.transform(limits)
# data space
out = scale.break_info(rangee)
# trans'd range
out['range'] = np.sort(trans.transform(out['range']))
if limits is None:
expand = self.expand_default(scale)
out['range'] = expand_range_distinct(out['range'], expand)
# major and minor breaks in plot space
out['major'] = transform_value(trans, out['major'], out['range'])
out['minor'] = transform_value(trans, out['minor'], out['range'])
for key in list(out.keys()):
new_key = '{}_{}'.format(name, key)
out[new_key] = out.pop(key)
return out
def train(scale, limits, trans, name):
"""
Train a single coordinate axis
"""
if limits is None:
rangee = scale.dimension()
else:
rangee = scale.transform(limits)
# data space
out = scale.break_info(rangee)
# trans'd range
out['range'] = trans.transform(out['range'])
if limits is None:
expand = self.expand_default(scale)
out['range'] = expand_range_distinct(out['range'], expand)
# major and minor breaks in plot space
out['major'] = transform_value(trans, out['major'], out['range'])
out['minor'] = transform_value(trans, out['minor'], out['range'])
for key in list(out.keys()):
new_key = '{}_{}'.format(name, key)
out[new_key] = out.pop(key)
return out
def dimension(self, expand=(0, 0, 0, 0), limits=None):
"""
The phyical size of the scale, if a position scale
Unlike limits, this always returns a numeric vector of length 2
"""
if limits is None:
limits = self.limits
return expand_range_distinct(limits, expand)
def test_expand_range_distinct():
assert expand_range_distinct((0, 1)) == (0, 1)
assert expand_range_distinct((0, 1), (2, 0)) == (-2, 3)
assert expand_range_distinct((0, 1), (2, 0, 2, 0)) == (-2, 3)
assert expand_range_distinct((0, 1), (0, 2)) == (-2, 3)
assert expand_range_distinct((0, 1), (0, 2, 0, 2)) == (-2, 3)
assert expand_range_distinct((0, 1), (2, 2, 2, 2)) == (-4, 5)
assert expand_range_distinct((1, 1), (2, 2), zero_width=1) == (0.5, 1.5)
def test_expand_range_distinct():
assert expand_range_distinct((0, 1)) == (0, 1)
assert expand_range_distinct((0, 1), (2, 0)) == (-2, 3)
assert expand_range_distinct((0, 1), (2, 0, 2, 0)) == (-2, 3)
assert expand_range_distinct((0, 1), (0, 2)) == (-2, 3)
assert expand_range_distinct((0, 1), (0, 2, 0, 2)) == (-2, 3)
assert expand_range_distinct((0, 1), (2, 2, 2, 2)) == (-4, 5)
assert expand_range_distinct((1, 1), (2, 2), zero_width=1) == (0.5, 1.5)
def dimension(self, expand=(0, 0, 0, 0)):
"""
The phyical size of the scale, if a position scale
Unlike limits, this always returns a numeric vector of length 2
"""
c_range = self.range_c.range
d_range = self.limits
if self.is_empty():
return (0, 1)
elif self.range.range is None: # only continuous
return expand_range_distinct(c_range, expand)
elif c_range is None: # only discrete
# FIXME: I think this branch should not exist
return expand_range_distinct((1, len(d_range)), expand)
else: # both
# e.g categorical bar plot have discrete items, but
# are plot on a continuous x scale
a = np.hstack([
c_range,
expand_range_distinct((1, len(d_range)), expand)
])
return a.min(), a.max()
