Skip to content

Distplot fix #473

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 9 commits into from
May 27, 2016
Merged

Distplot fix #473

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
c2c1f18
changed histnorm to prob-density and adjusted both types of curves
yankev May 20, 2016
ada8147
fixed tests to comply with change to histnorm (-> probability density)
yankev May 20, 2016
1245b10
added line back at the end of the file
yankev May 20, 2016
b28b4a2
added histnorm parameter
yankev May 20, 2016
e4e53ef
added tests to cover both options of histnorm
yankev May 25, 2016
342fa54
edited docstring
yankev May 25, 2016
4570e59
changed default to probability density
yankev May 26, 2016
141a789
changed default value in docstring
yankev May 26, 2016
bda2531
put histnorm values into variables
yankev May 26, 2016
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
186 changes: 172 additions & 14 deletions plotly/tests/test_optional/test_figure_factory.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,13 +44,60 @@ def test_unequal_data_label_length(self):
self.assertRaises(PlotlyError, tls.FigureFactory.create_distplot,
**kwargs)

def test_simple_distplot(self):
def test_simple_distplot_prob_density(self):

# we should be able to create a single distplot with a simple dataset
# and default kwargs

dp = tls.FigureFactory.create_distplot(hist_data=[[1, 2, 2, 3]],
group_labels=['distplot'])
group_labels=['distplot'],
histnorm='probability density')
expected_dp_layout = {'barmode': 'overlay',
'hovermode': 'closest',
'legend': {'traceorder': 'reversed'},
'xaxis1': {'anchor': 'y2', 'domain': [0.0, 1.0], 'zeroline': False},
'yaxis1': {'anchor': 'free', 'domain': [0.35, 1], 'position': 0.0},
'yaxis2': {'anchor': 'x1',
'domain': [0, 0.25],
'dtick': 1,
'showticklabels': False}}
self.assertEqual(dp['layout'], expected_dp_layout)

expected_dp_data_hist = {'autobinx': False,
'histnorm': 'probability density',
'legendgroup': 'distplot',
'marker': {'color': 'rgb(31, 119, 180)'},
'name': 'distplot',
'opacity': 0.7,
'type': 'histogram',
'x': [1, 2, 2, 3],
'xaxis': 'x1',
'xbins': {'end': 3.0, 'size': 1.0, 'start': 1.0},
'yaxis': 'y1'}
self.assertEqual(dp['data'][0], expected_dp_data_hist)

expected_dp_data_rug = {'legendgroup': 'distplot',
'marker': {'color': 'rgb(31, 119, 180)',
'symbol': 'line-ns-open'},
'mode': 'markers',
'name': 'distplot',
'showlegend': False,
'text': None,
'type': 'scatter',
'x': [1, 2, 2, 3],
'xaxis': 'x1',
'y': ['distplot', 'distplot',
'distplot', 'distplot'],
'yaxis': 'y2'}
self.assertEqual(dp['data'][2], expected_dp_data_rug)

def test_simple_distplot_prob(self):

# we should be able to create a single distplot with a simple dataset
# and default kwargs

dp = tls.FigureFactory.create_distplot(hist_data=[[1, 2, 2, 3]],
group_labels=['distplot'], histnorm='probability')
expected_dp_layout = {'barmode': 'overlay',
'hovermode': 'closest',
'legend': {'traceorder': 'reversed'},
Expand Down Expand Up @@ -90,7 +137,69 @@ def test_simple_distplot(self):
'yaxis': 'y2'}
self.assertEqual(dp['data'][2], expected_dp_data_rug)

def test_distplot_more_args(self):
def test_distplot_more_args_prob_dens(self):

# we should be able to create a distplot with 2 datasets no
# rugplot, defined bin_size, and added title

hist1_x = [0.8, 1.2, 0.2, 0.6, 1.6,
-0.9, -0.07, 1.95, 0.9, -0.2,
-0.5, 0.3, 0.4, -0.37, 0.6]
hist2_x = [0.8, 1.5, 1.5, 0.6, 0.59,
1.0, 0.8, 1.7, 0.5, 0.8,
-0.3, 1.2, 0.56, 0.3, 2.2]

hist_data = [hist1_x] + [hist2_x]
group_labels = ['2012', '2013']

dp = tls.FigureFactory.create_distplot(hist_data, group_labels,
histnorm='probability density',
show_rug=False, bin_size=.2)
dp['layout'].update(title='Dist Plot')

expected_dp_layout = {'barmode': 'overlay',
'hovermode': 'closest',
'legend': {'traceorder': 'reversed'},
'title': 'Dist Plot',
'xaxis1': {'anchor': 'y2', 'domain': [0.0, 1.0],
'zeroline': False},
'yaxis1': {'anchor': 'free', 'domain': [0.0, 1],
'position': 0.0}}
self.assertEqual(dp['layout'], expected_dp_layout)

expected_dp_data_hist_1 = {'autobinx': False,
'histnorm': 'probability density',
'legendgroup': '2012',
'marker': {'color': 'rgb(31, 119, 180)'},
'name': '2012',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.2, 0.2, 0.6, 1.6, -0.9, -0.07,
1.95, 0.9, -0.2, -0.5, 0.3, 0.4,
-0.37, 0.6],
'xaxis': 'x1',
'xbins': {'end': 1.95, 'size': 0.2,
'start': -0.9},
'yaxis': 'y1'}
self.assertEqual(dp['data'][0], expected_dp_data_hist_1)

expected_dp_data_hist_2 = {'autobinx': False,
'histnorm': 'probability density',
'legendgroup': '2013',
'marker': {'color': 'rgb(255, 127, 14)'},
'name': '2013',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.5, 1.5, 0.6, 0.59, 1.0, 0.8,
1.7, 0.5, 0.8, -0.3, 1.2, 0.56, 0.3,
2.2],
'xaxis': 'x1',
'xbins': {'end': 2.2, 'size': 0.2,
'start': -0.3},
'yaxis': 'y1'}
self.assertEqual(dp['data'][1], expected_dp_data_hist_2)

def test_distplot_more_args_prob(self):

# we should be able to create a distplot with 2 datasets no
# rugplot, defined bin_size, and added title
Expand Down Expand Up @@ -151,7 +260,54 @@ def test_distplot_more_args(self):
'yaxis': 'y1'}
self.assertEqual(dp['data'][1], expected_dp_data_hist_2)

def test_distplot_binsize_array(self):
def test_distplot_binsize_array_prob(self):
hist1_x = [0.8, 1.2, 0.2, 0.6, 1.6, -0.9, -0.07, 1.95, 0.9, -0.2,
-0.5, 0.3, 0.4, -0.37, 0.6]
hist2_x = [0.8, 1.5, 1.5, 0.6, 0.59, 1.0, 0.8, 1.7, 0.5, 0.8, -0.3,
1.2, 0.56, 0.3, 2.2]

hist_data = [hist1_x, hist2_x]
group_labels = ['2012', '2013']

dp = tls.FigureFactory.create_distplot(hist_data, group_labels,
show_rug=False,
bin_size=[.2, .2])

expected_dp_data_hist_1 = {'autobinx': False,
'histnorm': 'probability density',
'legendgroup': '2012',
'marker':
{'color': 'rgb(31, 119, 180)'},
'name': '2012',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.2, 0.2, 0.6, 1.6, -0.9,
-0.07, 1.95, 0.9, -0.2, -0.5, 0.3,
0.4, -0.37, 0.6],
'xaxis': 'x1',
'xbins': {'end': 1.95, 'size': 0.2,
'start': -0.9},
'yaxis': 'y1'}
self.assertEqual(dp['data'][0], expected_dp_data_hist_1)

expected_dp_data_hist_2 = {'autobinx': False,
'histnorm': 'probability density',
'legendgroup': '2013',
'marker':
{'color': 'rgb(255, 127, 14)'},
'name': '2013',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.5, 1.5, 0.6, 0.59, 1.0,
0.8, 1.7, 0.5, 0.8, -0.3, 1.2,
0.56, 0.3, 2.2],
'xaxis': 'x1',
'xbins': {'end': 2.2, 'size': 0.2,
'start': -0.3},
'yaxis': 'y1'}
self.assertEqual(dp['data'][1], expected_dp_data_hist_2)

def test_distplot_binsize_array_prob_density(self):
hist1_x = [0.8, 1.2, 0.2, 0.6, 1.6, -0.9, -0.07, 1.95, 0.9, -0.2,
-0.5, 0.3, 0.4, -0.37, 0.6]
hist2_x = [0.8, 1.5, 1.5, 0.6, 0.59, 1.0, 0.8, 1.7, 0.5, 0.8, -0.3,
Expand All @@ -165,31 +321,33 @@ def test_distplot_binsize_array(self):
bin_size=[.2, .2])

expected_dp_data_hist_1 = {'autobinx': False,
'histnorm': 'probability',
'histnorm': 'probability density',
'legendgroup': '2012',
'marker': {'color': 'rgb(31, 119, 180)'},
'marker':
{'color': 'rgb(31, 119, 180)'},
'name': '2012',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.2, 0.2, 0.6, 1.6, -0.9, -0.07,
1.95, 0.9, -0.2, -0.5, 0.3, 0.4,
-0.37, 0.6],
'x': [0.8, 1.2, 0.2, 0.6, 1.6, -0.9,
-0.07, 1.95, 0.9, -0.2, -0.5, 0.3,
0.4, -0.37, 0.6],
'xaxis': 'x1',
'xbins': {'end': 1.95, 'size': 0.2,
'start': -0.9},
'yaxis': 'y1'}
self.assertEqual(dp['data'][0], expected_dp_data_hist_1)

expected_dp_data_hist_2 = {'autobinx': False,
'histnorm': 'probability',
'histnorm': 'probability density',
'legendgroup': '2013',
'marker': {'color': 'rgb(255, 127, 14)'},
'marker':
{'color': 'rgb(255, 127, 14)'},
'name': '2013',
'opacity': 0.7,
'type': 'histogram',
'x': [0.8, 1.5, 1.5, 0.6, 0.59, 1.0, 0.8,
1.7, 0.5, 0.8, -0.3, 1.2, 0.56, 0.3,
2.2],
'x': [0.8, 1.5, 1.5, 0.6, 0.59, 1.0,
0.8, 1.7, 0.5, 0.8, -0.3, 1.2,
0.56, 0.3, 2.2],
'xaxis': 'x1',
'xbins': {'end': 2.2, 'size': 0.2,
'start': -0.3},
Expand Down
24 changes: 14 additions & 10 deletions plotly/tools.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4019,7 +4019,7 @@ def create_candlestick(open, high, low, close,
@staticmethod
def create_distplot(hist_data, group_labels,
bin_size=1., curve_type='kde',
colors=[], rug_text=[],
colors=[], rug_text=[], histnorm='probability',
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@yankev can you add that param to the docstring as well (so around line 4039)

show_hist=True, show_curve=True,
show_rug=True):
"""
Expand Down Expand Up @@ -4144,23 +4144,23 @@ def create_distplot(hist_data, group_labels,
bin_size = [bin_size]*len(hist_data)

hist = _Distplot(
hist_data, group_labels, bin_size,
hist_data, histnorm, group_labels, bin_size,
curve_type, colors, rug_text,
show_hist, show_curve).make_hist()

if curve_type == 'normal':
curve = _Distplot(
hist_data, group_labels, bin_size,
hist_data, histnorm, group_labels, bin_size,
curve_type, colors, rug_text,
show_hist, show_curve).make_normal()
else:
curve = _Distplot(
hist_data, group_labels, bin_size,
hist_data, histnorm, group_labels, bin_size,
curve_type, colors, rug_text,
show_hist, show_curve).make_kde()

rug = _Distplot(
hist_data, group_labels, bin_size,
hist_data, histnorm, group_labels, bin_size,
curve_type, colors, rug_text,
show_hist, show_curve).make_rug()

Expand Down Expand Up @@ -5036,10 +5036,11 @@ class _Distplot(FigureFactory):
"""
Refer to TraceFactory.create_distplot() for docstring
"""
def __init__(self, hist_data, group_labels,
def __init__(self, hist_data, histnorm, group_labels,
bin_size, curve_type, colors,
rug_text, show_hist, show_curve):
self.hist_data = hist_data
self.histnorm = histnorm
self.group_labels = group_labels
self.bin_size = bin_size
self.show_hist = show_hist
Expand Down Expand Up @@ -5081,7 +5082,7 @@ def make_hist(self):
x=self.hist_data[index],
xaxis='x1',
yaxis='y1',
histnorm='probability',
histnorm=self.histnorm,
name=self.group_labels[index],
legendgroup=self.group_labels[index],
marker=dict(color=self.colors[index]),
Expand All @@ -5108,7 +5109,9 @@ def make_kde(self):
self.curve_y[index] = (scipy.stats.gaussian_kde
(self.hist_data[index])
(self.curve_x[index]))
self.curve_y[index] *= self.bin_size[index]

if self.histnorm == 'probability':
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you pull these into constants at the top of the file? Hard-coded things like this always raise 🚨's for me.

self.curve_y[index] *= self.bin_size[index]

for index in range(self.trace_number):
curve[index] = dict(type='scatter',
Expand Down Expand Up @@ -5143,7 +5146,9 @@ def make_normal(self):
/ 500 for x in range(500)]
self.curve_y[index] = scipy.stats.norm.pdf(
self.curve_x[index], loc=mean[index], scale=sd[index])
self.curve_y[index] *= self.bin_size[index]

if self.histnorm == 'probability':
self.curve_y[index] *= self.bin_size[index]

for index in range(self.trace_number):
curve[index] = dict(type='scatter',
Expand Down Expand Up @@ -5621,4 +5626,3 @@ def make_table_annotations(self):
font=dict(color=font_color),
showarrow=False))
return annotations