def test_acceptance(validator: CompoundValidator):
val = Marker(color='green', size=10)
res = validator.validate_coerce(val)
assert isinstance(res, Marker)
assert res.color == 'green'
assert res.size == 10
def plot_residuals(self, data, title):
''' Plots the residuals of the predicted values
Parameters
==========
data: pandas dataframe
Dataframe that contains the actual (y) and predicted (y_pred) values.
'''
data.iplot(data=[
dict(y=data['y'] - data['y_pred'],
x=data['Dates'],
name='Actual',
mode='markers',
marker=Marker(color='Black', size=3, opacity=1),
color='Gray',
showlegend=False),
],
title=title,
dimensions=(700, 400),
vspan=dict(x0=datetime.datetime.strptime(
self.testing_cutoff, "%Y%m%d"),
x1=data['Dates'].max(),
color='Black',
width=0,
fill=False,
opacity=0.1))
def test_acceptance(validator):
val = Marker(color="green", size=10)
res = validator.validate_coerce(val)
assert isinstance(res, Marker)
assert res.color == "green"
assert res.size == 10
def make_marker(obs, index):
"""
:param
obs: Observation point containing 'value', 'index' and 'name'
color: Same color as belonging line
:returns: a marker for the value of the given observation.
"""
return Scatter(y=[obs['value']],
x=[index],
showlegend=False,
mode='markers',
marker=Marker(color='#000', ),
hoverlabel=Hoverlabel(bgcolor='#000'))
def compare_portfolios(request):
if request.method == "POST":
portfolios = []
date_from = timezone.datetime(1999, 1, 1).date()
for portfolio in Portfolio.objects.all():
if request.POST.get(str(portfolio.id)) is not None:
portfolios.append(portfolio)
delta = date_from - portfolio.creation_date
if delta.days < 0:
date_from = portfolio.creation_date
stocks_list, min_create_date = get_all_tickers(portfolios), date_from
download_stocks(stocks_list, min_create_date)
R = [[
portfolio.name,
PortfolioHandler(portfolio).getRByDates(date_from)
] for portfolio in portfolios]
fig = go.Figure()
for name, r in R:
scatter = go.Scatter(x=r[0],
y=r[1],
mode='markers+lines',
marker=Marker(symbol='0'),
opacity=0.8,
name=name)
fig.add_trace(scatter)
plt_div = plot(fig, output_type='div')
return render(
request, 'pages/compare.html',
get_full_context(
request, {
'page': COMPARE_PORTFOLIOS_PAGE_NAME,
'portfolios': Portfolio.objects.all(),
'html': plt_div
}))
return render(
request, 'pages/compare.html',
get_full_context(
request, {
'page': COMPARE_PORTFOLIOS_PAGE_NAME,
'portfolios': Portfolio.objects.all(),
}))
def plot_act_exp(self, data, title):
''' Plots the actuals vs the expected values
Parameters
==========
data: pandas dataframe
Dataframe that contains the actual (y) and
the predicted (y_pred) values with the standard deviation y_std
'''
data.iplot(data=[
dict(y=data['y_pred'] - 1.96 * data['y_std'],
x=data['Dates'],
name='Predicted - 1.96 σ',
line=Line(color='Gray', width=0),
showlegend=False),
dict(y=data['y_pred'] + 1.96 * data['y_std'],
x=data['Dates'],
name='Predicted + 1.96 σ',
fill='tonexty',
line=Line(color='Gray', width=0),
showlegend=False),
dict(y=data['y'],
x=data['Dates'],
name='Actual',
mode='markers',
marker=Marker(color='Black', size=3, opacity=1),
color='Gray',
showlegend=False),
dict(y=data['y_pred'],
x=data['Dates'],
name='Predicted',
line=Line(color='Blue', width=0.8),
showlegend=False,
opacity=0.5),
],
title=title,
dimensions=(700, 400),
vspan=dict(x0=datetime.datetime.strptime(
self.testing_cutoff, "%Y%m%d"),
x1=data['Dates'].max(),
color='Black',
width=0,
fill=False,
opacity=0.1))
def get_single_propagation_scatter(self,
color: Optional[str] = None) -> Scatter:
time = defaultdict(int)
for child in self.all_children:
time[child.delay] += 1
x = sorted(time.keys())
y = []
for delay in x:
prev_sum = y[-1] if y else 0
y.append(prev_sum + time[delay])
return Scatter(x=x,
y=y,
marker=Marker(color=color) if color else None,
text=self.content)
def create_line_trace(net,
lines=None,
use_line_geodata=True,
respect_switches=False,
width=1.0,
color='grey',
infofunc=None,
trace_name='lines',
legendgroup=None,
cmap=None,
cbar_title=None,
show_colorbar=True,
cmap_vals=None,
cmin=None,
cmax=None,
cpos=1.1):
"""
Creates a plotly trace of pandapower lines.
INPUT:
**net** (pandapowerNet) - The pandapower network
OPTIONAL:
**lines** (list, None) - The lines for which the collections are created.
If None, all lines in the network are considered.
**width** (int, 1) - line width
**respect_switches** (bool, False) - flag for consideration of disconnected lines
**infofunc** (pd.Series, None) - hoverinfo for line elements. Indices should correspond to
the pandapower element indices
**trace_name** (String, "lines") - name of the trace which will appear in the legend
**color** (String, "grey") - color of lines in the trace
**legendgroup** (String, None) - defines groups of layers that will be displayed in a legend
e.g. groups according to voltage level (as used in `vlevel_plotly`)
**cmap** (String, None) - name of a colormap which exists within plotly if set to True default `Jet`
colormap is used, alternative colormaps : Greys, YlGnBu, Greens, YlOrRd,
Bluered, RdBu, Reds, Blues, Picnic, Rainbow, Portland, Jet, Hot, Blackbody, Earth, Electric, Viridis
**cmap_vals** (list, None) - values used for coloring using colormap
**show_colorbar** (bool, False) - flag for showing or not corresponding colorbar
**cbar_title** (String, None) - title for the colorbar
**cmin** (float, None) - colorbar range minimum
**cmax** (float, None) - colorbar range maximum
**cpos** (float, 1.1) - position of the colorbar
"""
color = get_plotly_color(color)
# defining lines to be plot
lines = net.line.index.tolist() if lines is None else list(lines)
if len(lines) == 0:
return []
if infofunc is not None:
if not isinstance(infofunc, pd.Series) and isinstance(infofunc, Iterable) and \
len(infofunc) == len(lines):
infofunc = pd.Series(index=lines, data=infofunc)
if len(infofunc) != len(lines) and len(infofunc) != len(net.line):
raise UserWarning(
"Different amount of hover info than lines to plot")
assert isinstance(infofunc, pd.Series), \
"infofunc should be a pandas series with the net.line.index to the infofunc contents"
no_go_lines = set()
if respect_switches:
no_go_lines = set(lines) & set(net.switch.element[
(net.switch.et == "l") & (net.switch.closed == 0)])
lines_to_plot = net.line.loc[set(net.line.index)
& (set(lines) - no_go_lines)]
no_go_lines_to_plot = None
use_line_geodata = use_line_geodata if net.line_geodata.shape[
0] > 0 else False
if use_line_geodata:
lines_to_plot = lines_to_plot.loc[set(lines_to_plot.index)
& set(net.line_geodata.index)]
else:
lines_with_geodata = lines_to_plot.from_bus.isin(net.bus_geodata.index) & \
lines_to_plot.to_bus.isin(net.bus_geodata.index)
lines_to_plot = lines_to_plot.loc[lines_with_geodata]
cmap_lines = None
if cmap is not None:
# workaround: if colormap plot is used, each line need to be separate scatter object because
# plotly still doesn't support appropriately colormap for line objects
# TODO correct this when plotly solves existing github issue about Line colorbar
cmap = 'jet' if cmap is True else cmap
if cmap_vals is not None:
if not isinstance(cmap_vals, np.ndarray):
cmap_vals = np.asarray(cmap_vals)
else:
if net.res_line.shape[0] == 0:
logger.error(
"There are no power flow results for lines which are default for line colormap coloring..."
"set cmap_vals input argument if you want colormap according to some specific values..."
)
cmap_vals = net.res_line.loc[lines_to_plot.index,
'loading_percent'].values
cmap_lines = get_plotly_cmap(cmap_vals,
cmap_name=cmap,
cmin=cmin,
cmax=cmax)
if len(cmap_lines) == len(net.line):
# some lines are not plotted although cmap_value were provided for all lines
line_idx_map = dict(
zip(net.line.loc[lines].index.tolist(), range(len(lines))))
cmap_lines = [
cmap_lines[line_idx_map[idx]] for idx in lines_to_plot.index
]
else:
assert len(cmap_lines) == len(lines_to_plot), \
"Different amounts of cmap values and lines to plot were supplied"
line_traces = []
for col_i, (idx, line) in enumerate(lines_to_plot.iterrows()):
line_color = color
line_info = line['name']
if cmap is not None:
try:
line_color = cmap_lines[col_i]
line_info = line['name'] if infofunc is None else infofunc.loc[
idx]
except IndexError:
logger.warning(
"No color and info for line {:d} (name: {}) available".
format(idx, line['name']))
line_trace = dict(type='scatter',
text=[],
hoverinfo='text',
mode='lines',
name=trace_name,
line=Line(width=width, color=color))
line_trace['x'], line_trace['y'] = _get_line_geodata_plotly(
net, lines_to_plot.loc[idx:idx], use_line_geodata)
line_trace['line']['color'] = line_color
line_trace['text'] = line_info
line_traces.append(line_trace)
if show_colorbar and cmap is not None:
cmin = cmap_vals.min() if cmin is None else cmin
cmax = cmap_vals.max() if cmax is None else cmax
try:
# TODO for custom colormaps
cbar_cmap_name = 'Jet' if cmap == 'jet' else cmap
# workaround to get colorbar for lines (an unvisible node is added)
# get x and y of first line.from_bus:
x = [net.bus_geodata.x[net.line.from_bus[net.line.index[0]]]]
y = [net.bus_geodata.y[net.line.from_bus[net.line.index[0]]]]
lines_cbar = dict(type='scatter',
x=x,
y=y,
mode='markers',
marker=Marker(
size=0,
cmin=cmin,
cmax=cmax,
color='rgb(255,255,255)',
opacity=0,
colorscale=cbar_cmap_name,
colorbar=ColorBar(thickness=10, x=cpos),
))
if cbar_title:
lines_cbar['marker']['colorbar']['title'] = cbar_title
lines_cbar['marker']['colorbar']['title']['side'] = 'right'
line_traces.append(lines_cbar)
except:
pass
if len(no_go_lines) > 0:
no_go_lines_to_plot = net.line.loc[no_go_lines]
for idx, line in no_go_lines_to_plot.iterrows():
line_color = color
line_trace = dict(type='scatter',
text=[],
hoverinfo='text',
mode='lines',
name='disconnected lines',
line=Line(width=width / 2,
color='grey',
dash='dot'))
line_trace['x'], line_trace['y'] = _get_line_geodata_plotly(
net, no_go_lines_to_plot.loc[idx:idx], use_line_geodata)
line_trace['line']['color'] = line_color
try:
line_trace['text'] = infofunc.loc[idx]
except (KeyError, IndexError, AttributeError):
line_trace["text"] = line['name']
line_traces.append(line_trace)
if legendgroup:
line_trace['legendgroup'] = legendgroup
# sort infofunc so that it is the correct order lines_to_plot + no_go_lines_to_plot
if infofunc is not None:
if not isinstance(infofunc, pd.Series) and isinstance(infofunc, Iterable) and \
len(infofunc) == len(net.line):
infofunc = pd.Series(index=net.line.index, data=infofunc)
assert isinstance(infofunc, pd.Series), \
"infofunc should be a pandas series with the net.line.index to the infofunc contents"
sorted_idx = lines_to_plot.index.tolist()
if no_go_lines_to_plot is not None:
sorted_idx += no_go_lines_to_plot.index.tolist()
infofunc = infofunc.loc[sorted_idx]
center_trace = create_edge_center_trace(line_traces,
color=color,
infofunc=infofunc,
use_line_geodata=use_line_geodata)
line_traces.append(center_trace)
return line_traces
def create_bus_trace(net,
buses=None,
size=5,
patch_type="circle",
color="blue",
infofunc=None,
trace_name='buses',
legendgroup=None,
cmap=None,
cmap_vals=None,
cbar_title=None,
cmin=None,
cmax=None,
cpos=1.0,
colormap_column="vm_pu"):
"""
Creates a plotly trace of pandapower buses.
INPUT:
**net** (pandapowerNet) - The pandapower network
OPTIONAL:
**buses** (list, None) - The buses for which the collections are created.
If None, all buses in the network are considered.
**size** (int, 5) - patch size
**patch_type** (str, "circle") - patch type, can be
- "circle" for a circle
- "square" for a rectangle
- "diamond" for a diamond
- much more pathc types at https://plot.ly/python/reference/#scatter-marker
**infofunc** (pd.Series, None) - hoverinfo for bus elements. Indices should correspond to
the pandapower element indices
**trace_name** (String, "buses") - name of the trace which will appear in the legend
**color** (String, "blue") - color of buses in the trace
**cmap** (String, None) - name of a colormap which exists within plotly
(Greys, YlGnBu, Greens, YlOrRd, Bluered, RdBu, Reds, Blues, Picnic, Rainbow,
Portland, Jet, Hot, Blackbody, Earth, Electric, Viridis) alternatively a custom
discrete colormap can be used
**cmap_vals** (list, None) - values used for coloring using colormap
**cbar_title** (String, None) - title for the colorbar
**cmin** (float, None) - colorbar range minimum
**cmax** (float, None) - colorbar range maximum
**cpos** (float, 1.1) - position of the colorbar
**colormap_column** (str, "vm_pu") - set color of bus according to this variable
"""
color = get_plotly_color(color)
bus_trace = dict(type='scatter',
text=[],
mode='markers',
hoverinfo='text',
name=trace_name,
marker=dict(color=color, size=size, symbol=patch_type))
buses = net.bus.index.tolist() if buses is None else list(buses)
bus_plot_index = [
b for b in buses if b in list(set(buses) & set(net.bus_geodata.index))
]
bus_trace['x'], bus_trace['y'] = (net.bus_geodata.loc[bus_plot_index,
'x'].tolist(),
net.bus_geodata.loc[bus_plot_index,
'y'].tolist())
if not isinstance(infofunc, pd.Series) and isinstance(infofunc, Iterable) and \
len(infofunc) == len(buses):
infofunc = pd.Series(index=buses, data=infofunc)
bus_trace['text'] = net.bus.loc[bus_plot_index, 'name'] if infofunc is None else \
infofunc.loc[buses]
if legendgroup:
bus_trace['legendgroup'] = legendgroup
# if color map is set
if cmap is not None:
# TODO introduce discrete colormaps (see contour plots in plotly)
# if cmap_vals are not given
cmap = 'Jet' if cmap is True else cmap
if cmap_vals is not None:
cmap_vals = cmap_vals
else:
if net.res_line.shape[0] == 0:
logger.error(
"There are no power flow results for buses voltage magnitudes which are default for bus "
"colormap coloring..."
"set cmap_vals input argument if you want colormap according to some specific values..."
)
cmap_vals = net.res_bus.loc[bus_plot_index, colormap_column].values
cmap_vals = net.res_bus.loc[
bus_plot_index,
colormap_column] if cmap_vals is None else cmap_vals
cmin = cmap_vals.min() if cmin is None else cmin
cmax = cmap_vals.max() if cmax is None else cmax
bus_trace['marker'] = Marker(size=size,
color=cmap_vals,
cmin=cmin,
cmax=cmax,
colorscale=cmap,
colorbar=ColorBar(thickness=10, x=cpos),
symbol=patch_type)
if cbar_title:
bus_trace['marker']['colorbar']['title'] = cbar_title
bus_trace['marker']['colorbar']['title']['side'] = 'right'
return [bus_trace]
def _create_branch_trace(net,
branches=None,
use_branch_geodata=True,
respect_separators=False,
width=1.0,
color='grey',
infofunc=None,
trace_name='lines',
legendgroup=None,
cmap=None,
cbar_title=None,
show_colorbar=True,
cmap_vals=None,
cmin=None,
cmax=None,
cpos=1.1,
branch_element='line',
separator_element='switch',
node_element='bus',
cmap_vals_category='loading_percent'):
"""
Creates a plotly trace of branch elements. The rather generic, non-power net specific names
were introduced to make it usable in other packages, e.g. for pipe networks.
INPUT:
**net** (pandapowerNet) - The network
OPTIONAL:
**branches** (list, None) - The branches for which the collections are created.
If None, all branches in the network are considered.
**use_branch_geodata** (bool, True) - whether the geodata of the branch tables should be used
**respect_separators** (bool, True) - whether separating elements like switches should be
considered
**width** (int, 1) - branch width
**color** (String, "grey") - color of lines in the trace
**infofunc** (pd.Series, None) - hoverinfo for line elements. Indices should correspond to
the pandapower element indices
**trace_name** (String, "lines") - name of the trace which will appear in the legend
**legendgroup** (String, None) - defines groups of layers that will be displayed in a legend
e.g. groups according to voltage level (as used in `vlevel_plotly`)
**cmap** (String, None) - name of a colormap which exists within plotly if set to True default `Jet`
colormap is used, alternative colormaps : Greys, YlGnBu, Greens, YlOrRd,
Bluered, RdBu, Reds, Blues, Picnic, Rainbow, Portland, Jet, Hot, Blackbody, Earth, Electric, Viridis
**cmap_vals** (list, None) - values used for coloring using colormap
**show_colorbar** (bool, False) - flag for showing or not corresponding colorbar
**cbar_title** (String, None) - title for the colorbar
**cmin** (float, None) - colorbar range minimum
**cmax** (float, None) - colorbar range maximum
**cpos** (float, 1.1) - position of the colorbar
**branch_element** (str, "line") - name of the branch element in the net. In a pandapower
net, this is alwas "line"
**separator_element** (str, "switch") - name of the separator element in the net. In a
pandapower net, this is alwas "switch"
**node_element** (str, "bus") - name of the node element in the net. In a pandapower net,
this is alwas "bus" (net.bus)
"""
color = get_plotly_color(color)
# defining branches (lines) to be plot
branches = net[branch_element].index.tolist(
) if branches is None else list(branches)
if len(branches) == 0:
return []
if infofunc is not None:
if not isinstance(infofunc, pd.Series) and isinstance(infofunc, Iterable) and \
len(infofunc) == len(branches):
infofunc = pd.Series(index=branches, data=infofunc)
if len(infofunc) != len(branches) and len(infofunc) != len(
net[branch_element]):
raise UserWarning("Different amount of hover info than {}s to "
"plot".format(branch_element))
assert isinstance(infofunc, pd.Series), \
"infofunc should be a pandas series with the net.{}.index to the infofunc " \
"contents".format(branch_element)
no_go_branches = set()
if respect_separators:
if separator_element == "switch":
no_go_branches = set(branches) & \
set(net[separator_element].element[(net[separator_element].et == "l") &
(net[separator_element].closed == 0)])
elif separator_element == "valve":
no_go_branches = set(branches) & \
set(net[separator_element][(~net[separator_element].in_service) |
(net[separator_element].opened)])
else:
raise NotImplementedError(
"respect separtors is only implements for switches, "
"not for {}s.".format(separator_element))
branches_to_plot = net[branch_element].loc[set(net[branch_element].index)
& (set(branches) -
no_go_branches)]
no_go_branches_to_plot = None
branch_geodata = branch_element + "_geodata"
node_geodata = node_element + "_geodata"
use_branch_geodata = use_branch_geodata if net[branch_geodata].shape[
0] > 0 else False
if use_branch_geodata:
branches_to_plot = branches_to_plot.loc[
set(branches_to_plot.index) & set(net[branch_geodata].index)]
else:
branches_with_geodata = branches_to_plot['from_'+node_element].isin(
net[node_geodata].index) & \
branches_to_plot['to_'+node_element].isin(net[node_geodata].index)
branches_to_plot = branches_to_plot.loc[branches_with_geodata]
cmap_branches = None
if cmap is not None:
# workaround: if colormap plot is used, each line need to be separate scatter object because
# plotly still doesn't support appropriately colormap for line objects
# TODO correct this when plotly solves existing github issue about Line colorbar
cmap = 'jet' if cmap is True else cmap
if cmap_vals is not None:
if not isinstance(cmap_vals, np.ndarray):
cmap_vals = np.asarray(cmap_vals)
else:
if net['res_' + branch_element].shape[0] == 0:
logger.error(
"There are no simulation results for branches which are default for {}"
"colormap coloring..."
"set cmap_vals input argument if you want colormap according to some specific "
"values...".format(branch_element))
cmap_vals = net['res_' +
branch_element].loc[branches_to_plot.index,
cmap_vals_category].values
cmap_branches = get_plotly_cmap(cmap_vals,
cmap_name=cmap,
cmin=cmin,
cmax=cmax)
if len(cmap_branches) == len(net[branch_element]):
# some branches are not plotted although cmap_value were provided for all branches
branch_idx_map = dict(
zip(net[branch_element].loc[branches].index.tolist(),
range(len(branches))))
cmap_branches = [
cmap_branches[branch_idx_map[idx]]
for idx in branches_to_plot.index
]
else:
assert len(cmap_branches) == len(branches_to_plot), \
"Different amounts of cmap values and branches to plot were supplied"
branch_traces = []
for col_i, (idx, branch) in enumerate(branches_to_plot.iterrows()):
line_color = color
line_info = branch['name']
if cmap is not None:
try:
line_color = cmap_branches[col_i]
line_info = branch[
'name'] if infofunc is None else infofunc.loc[idx]
except IndexError:
logger.warning(
"No color and info for {} {:d} (name: {}) available".
format(branch_element, idx, branch['name']))
line_trace = dict(type='scatter',
text=[],
hoverinfo='text',
mode='lines',
name=trace_name,
line=Line(width=width, color=color))
line_trace['x'], line_trace['y'] = _get_branch_geodata_plotly(
net, branches_to_plot.loc[idx:idx], use_branch_geodata,
branch_element, node_element)
line_trace['line']['color'] = line_color
line_trace['text'] = line_info
branch_traces.append(line_trace)
if show_colorbar and cmap is not None:
cmin = cmap_vals.min() if cmin is None else cmin
cmax = cmap_vals.max() if cmax is None else cmax
try:
# TODO for custom colormaps
cbar_cmap_name = 'Jet' if cmap == 'jet' else cmap
# workaround to get colorbar for branches (an unvisible node is added)
# get x and y of first line.from_bus:
x = [
net[node_geodata].x[net[branch_element]["from_" + node_element]
[net[branch_element].index[0]]]
]
y = [
net[node_geodata].y[net[branch_element]["from_" + node_element]
[net[branch_element].index[0]]]
]
branches_cbar = dict(type='scatter',
x=x,
y=y,
mode='markers',
marker=Marker(
size=0,
cmin=cmin,
cmax=cmax,
color='rgb(255,255,255)',
opacity=0,
colorscale=cbar_cmap_name,
colorbar=ColorBar(thickness=10, x=cpos),
))
if cbar_title:
branches_cbar['marker']['colorbar']['title'] = cbar_title
branches_cbar['marker']['colorbar']['title']['side'] = 'right'
branch_traces.append(branches_cbar)
except:
pass
if len(no_go_branches) > 0:
no_go_branches_to_plot = net[branch_element].loc[no_go_branches]
for idx, branch in no_go_branches_to_plot.iterrows():
line_color = color
line_trace = dict(type='scatter',
text=[],
hoverinfo='text',
mode='lines',
name='disconnected branches',
line=Line(width=width / 2,
color='grey',
dash='dot'))
line_trace['x'], line_trace['y'] = _get_branch_geodata_plotly(
net, no_go_branches_to_plot.loc[idx:idx], use_branch_geodata,
branch_element, node_element)
line_trace['line']['color'] = line_color
try:
line_trace['text'] = infofunc.loc[idx]
except (KeyError, IndexError, AttributeError):
line_trace["text"] = branch['name']
branch_traces.append(line_trace)
if legendgroup:
line_trace['legendgroup'] = legendgroup
# sort infofunc so that it is the correct order lines_to_plot + no_go_lines_to_plot
if infofunc is not None:
if not isinstance(infofunc, pd.Series) and isinstance(infofunc, Iterable) and \
len(infofunc) == len(net[branch_element]):
infofunc = pd.Series(index=net[branch_element].index,
data=infofunc)
assert isinstance(infofunc, pd.Series), \
"infofunc should be a pandas series with the net.{}.index to the infofunc contents" \
.format(branch_element)
sorted_idx = branches_to_plot.index.tolist()
if no_go_branches_to_plot is not None:
sorted_idx += no_go_branches_to_plot.index.tolist()
infofunc = infofunc.loc[sorted_idx]
center_trace = create_edge_center_trace(
branch_traces,
color=color,
infofunc=infofunc,
use_line_geodata=use_branch_geodata)
branch_traces.append(center_trace)
return branch_traces
def offline_plotly_scatter_bubble(
df,
x='x',
y='y',
size_col='size',
text_col='text',
category_col='category',
possible_categories=None,
filename=None,
config={'displaylogo': False},
xscale=None,
yscale='log',
layout={
'hovermode': 'closest',
'showlegend': False,
'autosize': True
},
marker={'sizemode': 'area'},
min_size=10,
):
r"""Interactive scatterplot of a DataFrame with the size and color of circles linke to two columns
config keys:
fillFrame setBackground displaylogo sendData showLink linkText staticPlot scrollZoom plot3dPixelRatio displayModeBar
showTips workspace doubleClick autosizable editable
layout keys:
angularaxis annotations autosize bargap bargroupgap barmode barnorm boxgap boxgroupgap boxmode calendar
direction dragmode font geo height hiddenlabels hiddenlabelssrc hidesources hovermode images legend
mapbox margin orientation paper_bgcolor plot_bgcolor radialaxis scene separators shapes showlegend sliders smith
ternary title titlefont updatemenus width xaxis yaxis
marker keys:
autocolorscale blend border cauto cmax cmin color colorbar colors colorscale colorsrc colorssrc line maxdisplayed
opacity opacitysrc outliercolor reversescale showscale size sizemax sizemin sizemode sizeref sizesrc symbol symbolsrc
marker['sizeref'] gives the denominator of the circle scaling factor.
Typically it should be about a tenth of the minimum 'size' column value
>>> from nlpia.data.loaders import get_data
>>> df = get_data('cities_us_wordvectors_pca2_meta').iloc[:100]
>>> html = offline_plotly_scatter_bubble(
... df.sort_values('population', ascending=False)[:350].copy().sort_values('population'),
... x='x', y='y',
... size_col='population', text_col='name', category_col='timezone',
... xscale=None, yscale=None, # 'log' or None
... layout={}, marker={'sizeref': 3000})
"""
config_default = dict(DEFAULT_PLOTLY_CONFIG)
marker_default = {
'size': size_col or min_size,
'sizemode': 'area',
'sizeref': int(df[size_col].min() * .8) if size_col else min_size
}
marker_default.update(marker)
size_col = marker_default.pop('size')
layout_default = {
'xaxis': XAxis(title=x, type=xscale),
'yaxis': YAxis(title=y, type=yscale),
}
layout_default.update(**layout)
if config is not None:
config_default.update(config)
df.columns = clean_columns(df.columns)
if possible_categories is None and category_col is not None:
if category_col in df.columns:
category_labels = df[category_col]
else:
category_labels = np.array(category_col)
possible_categories = list(set(category_labels))
possible_categories = [
None
] if possible_categories is None else possible_categories
if category_col and category_col in df:
masks = [
np.array(df[category_col] == label)
for label in possible_categories
]
else:
masks = [np.array([True] * len(df))] * len(possible_categories)
data = {
'data': [
Scatter(x=df[x][mask].values,
y=df[y][mask].values,
text=df[text_col][mask].values,
marker=Marker(size=df[size_col][mask]
if size_col in df.columns else size_col,
**marker_default),
mode='markers',
name=str(category_name))
for (category_name, mask) in zip(possible_categories, masks)
],
'layout':
Layout(**layout_default)
}
return offline_plotly_data(data, filename=filename, config=config_default)
def make_graph(nodes: list, edges: list):
"""
Creates plotly network graph.
Parameters
----------
nodes: list of nodes compatible with networx
edges: list of edges compatible with networx
Returns
-------
plotly.graph_objs.Figure
"""
G = nx.Graph()
# Create graph
G.add_nodes_from(nodes)
G.add_edges_from(edges)
# Hack
edge_labels_x = []
edge_labels_y = []
edge_labels_txt = []
# Define appends for better performance
add_x = edge_labels_x.append
add_y = edge_labels_y.append
add_txt = edge_labels_txt.append
# Create edges plot
edge_traces = []
add = edge_traces.append
for edge in G.edges(data=True):
a, b, info = edge
x0, y0 = G.node[a]['pos']
x1, y1 = G.node[b]['pos']
trace = go.Scatter(x=[x0, x1, None],
y=[y0, y1, None],
line=_color_edge(info),
hoverinfo='none',
mode='lines')
add(trace)
add_x((x0 + x1) / 2)
add_y((y0 + y1) / 2)
add_txt(f"time: {info['time'] }")
# Hack for info hover on edges:
middle_node_trace = go.Scatter(x=edge_labels_x,
y=edge_labels_y,
text=edge_labels_txt,
mode='markers',
hoverinfo='text',
marker=Marker(opacity=0))
node_trace = go.Scatter(x=[],
y=[],
text=[],
mode='markers',
hoverinfo='text',
marker=Marker(color=[],
size=10,
line=dict(width=2)))
# Create nodes plot
for node in G.nodes():
the_node = G.node[node]
x, y = the_node['pos']
node_trace['x'] += tuple([x])
node_trace['y'] += tuple([y])
node_info = reduce(lambda a, b: f'{a} | {b}',
[f'{k} : {v}' for k, v in the_node.items()])
node_trace['text'] += tuple([node_info])
data = edge_traces + [node_trace, middle_node_trace]
return go.Figure(data=data,
layout=go.Layout(
showlegend=False,
hovermode='closest',
margin=dict(b=20, l=5, r=5, t=40),
annotations=[
dict(text='',
showarrow=False,
xref="paper",
yref="paper",
x=0.005,
y=-0.002)
],
xaxis=dict(showgrid=False,
zeroline=False,
showticklabels=False),
yaxis=dict(showgrid=False,
zeroline=False,
showticklabels=False,
scaleanchor="x",
scaleratio=1),
))