def plot_the_table(self, df1, colorscale=None):
dimension_list = list()
trace = None
# print(colorscale)
colorscale = [ # Let first 10% (0.1) of the values have color rgb(0, 0, 0)
[0, "rgb(223, 169, 102)"],
[0.1, "rgb(223, 169, 102)"],
# Let values between 10-20% of the min and max of z
# have color rgb(20, 20, 20)
[0.1, "rgb(147, 133, 64)"],
[0.2, "rgb(147, 133, 64)"],
# Values between 20-30% of the min and max of z
# have color rgb(40, 40, 40)
[0.2, "rgb(20, 17, 177)"],
[0.3, "rgb(20, 17, 177)"],
[0.3, "rgb(160, 17, 2)"],
[0.4, "rgb(160, 17, 2)"],
[0.4, "rgb(80, 80, 80)"],
[0.5, "rgb(80, 80, 80)"],
[0.5, "rgb(100, 100, 100)"],
[0.6, "rgb(100, 100, 100)"],
[0.6, "rgb(120, 120, 120)"],
[0.7, "rgb(120, 120, 120)"],
[0.7, "rgb(140, 140, 140)"],
[0.8, "rgb(140, 140, 140)"],
[0.8, "rgb(160, 160, 160)"],
[0.9, "rgb(160, 160, 160)"],
[0.9, "rgb(180, 180, 180)"],
[1.0, "rgb(180, 180, 180)"]
]
for value in df1.columns:
dimension_list.append(dict(label=value, values=df1[value]))
if colorscale == None:
trace = go.Parcoords(line=dict(color=df1[df1.columns[-1]]),
dimensions=dimension_list)
else:
trace = go.Parcoords(line=dict(color=df1[df1.columns[-1]],
colorscale=colorscale),
dimensions=dimension_list)
data = [trace]
layout = None
if len(df1.columns) > 35:
layout = go.Layout(plot_bgcolor='#E5E5E5',
paper_bgcolor='#E5E5E5',
dragmode='zoom',
width=8000,
height=1000)
else:
layout = go.Layout(plot_bgcolor='#E5E5E5',
paper_bgcolor='#E5E5E5',
dragmode='zoom',
height=1000)
return go.Figure(data=data, layout=layout)
def create_parallel_2(by_week_df_2):
session_id = str(uuid.uuid4())
activities_df, by_week_df = get_dataframe(session_id)
dimensions = []
for i in range(7):
dimensions.append(
dict(range=[0, 20],
label='{}'.format(days_dict[i]),
values=by_week_df_2['miles_{}'.format(i)]))
data = [
go.Parcoords(line=dict(color=by_week_df['miles'],
colorscale='Hot',
showscale=True,
reversescale=True),
opacity=0.5,
dimensions=dimensions,
hoverinfo='text')
]
layout = go.Layout(plot_bgcolor='#E5E5E5',
paper_bgcolor='#E5E5E5',
title='Miles per week broken down by day')
return go.Figure(data=data, layout=layout)
def update_parcoords(loa_range, boa_range, draft_range, velocity_range,
disp_range):
dff = df[(df['loa'] > loa_range[0]) & (df['loa'] < loa_range[1]) &
# (df['lwl'] > lwl_range[0]) & (df['lwl'] < lwl_range[1]) &
(df['boa'] > boa_range[0]) & (df['boa'] < boa_range[1]) &
# (df['bwl'] > bwl_range[0]) & (df['bwl'] < bwl_range[1]) &
(df['draft'] > draft_range[0]) & (df['draft'] < draft_range[1]) &
(df['velocity'] > velocity_range[0]) &
(df['velocity'] < velocity_range[1]) &
(df['vol_disp'] > disp_range[0]) &
(df['vol_disp'] < disp_range[1])]
return {
'data': [
go.Parcoords(dimensions=[
# dict(label='LWL', values=dff['lwl']),
dict(label='LOA', values=dff['loa']),
# dict(label='BWL', values=dff['bwl']),
dict(label='BOA', values=dff['boa']),
dict(label='Draft', values=dff['draft']),
dict(label='Vol. displacement', values=dff['vol_disp']),
dict(label='Resistance', values=dff['total_resistance'])
])
]
}
def chart_plot(by_week_df):
dimensions = []
for i in range(7):
dimensions.append(
dict(range=[0, 20],
constraintrange=[0, 20],
label='{}'.format(days_dict[i]),
values=by_week_df['miles_{}'.format(i)]))
data = [
go.Parcoords(line=dict(color=by_week_df['miles'],
colorscale='Hot',
showscale=True,
reversescale=True),
opacity=0.5,
dimensions=dimensions,
hoverinfo='text')
]
layout = go.Layout(plot_bgcolor='#E5E5E5',
paper_bgcolor='#E5E5E5',
title='Miles per week broken down by day')
fig = go.Figure(data=data, layout=layout)
return plotly.offline.plot(fig, include_plotlyjs=False, output_type='div')
def parallel_coordinates_window(vis,
dimensions,
title,
line_color='blue',
win=None):
data = go.Parcoords(line=dict(color=line_color), dimensions=dimensions)
return vis._send({'data': [data], 'layout': dict(title=title), 'win': win})
def setUp(self):
# Construct initial scatter object
self.parcoords = go.Parcoords(name='parcoords A')
# Assert initial state
d1, d2 = strip_dict_params(
self.parcoords,
{'type': 'parcoords',
'name': 'parcoords A'}
)
assert d1 == d2
# Construct expected results
self.expected_toplevel = {
'type': 'parcoords',
'name': 'parcoords A',
'dimensions': [
{'values': [2, 3, 1], 'visible': True},
{'values': [1, 2, 3], 'label': 'dim1'}]}
self.layout = go.Layout()
self.expected_layout1 = {
'updatemenus': [{},
{'font': {'family': 'courier'}}]
}
self.expected_layout2 = {
'updatemenus': [{},
{'buttons': [
{}, {}, {'method': 'restyle'}]}]
}
def calc_graph(self):
graph = []
# PUT THE PARETO CURVE INSIDE
data = self.process_generation_total_performance_pareto()
data = self.normalize_data(data, self.normalization, self.objectives)
data = data.sort_values(['GHG_sys_tonCO2'])
dimensions = list([
dict(label=label, values=data[field]) if field != 'individual_name'
else dict(ticktext=data[field],
label=label,
tickvals=list(range(data.shape[0])),
values=list(range(data.shape[0])))
for field, label in zip(self.analysis_fields, self.titles)
])
line = dict(color=data['Capex_total_sys_USD'],
colorscale='Jet',
showscale=True)
trace = go.Parcoords(line=line,
dimensions=dimensions,
labelfont=dict(size=10),
rangefont=dict(size=8),
tickfont=dict(size=10))
graph.append(trace)
return graph
def parcoordsPlot(self, simId, scenario, resultName, numVars):
if simId is None or not (scenario and resultName):
return ''
result = self.getOutValues(simId, scenario, resultName)
inputsDF = self.getParameterValues(simId)
inputsDF = inputsDF.iloc[
result.index] # select only trials for which we have results
corrDF = self.getCorrDF(simId, scenario, resultName)
varCount = min(numVars, inputsDF.shape[1])
varNames = corrDF.index[:varCount]
self.paraCoordsVars = list(varNames)
def appendDim(series, name, constraint=None):
d = dict(range=[round(min(series), 2),
round(max(series), 2)],
label=name,
values=series)
if constraint:
# constrain by selection from distribution plot
d['constraintrange'] = constraint
dimensions.append(d)
dimensions = []
for name in varNames:
appendDim(inputsDF[name], name)
selected = self.selectedResults
appendDim(result,
resultName,
constraint=[min(selected), max(selected)])
plotData = [
go.Parcoords(
line=dict(color=result,
colorscale='Jet',
showscale=True,
reversescale=False,
cmin=min(result),
cmax=max(result)),
dimensions=dimensions,
rangefont={'color':
'rgba(255,0,0,0)'}, # zero alpha => invisible
)
]
layout = dict(
autosize=True,
height=500,
margin=dict(l=75, r=40, b=35, t=35),
hovermode="closest",
title='',
showlegend=False,
legend=dict(font=dict(size=10), orientation='h'),
)
figure = dict(data=plotData, layout=layout)
return figure
def plot_parallel_coordinate(df):
sample = df.sample(frac=0.1, replace=True, random_state=1)
fig = go.Figure(data=go.Parcoords(line=dict(
color=sample['K_Cluster'],
colorscale=[[0, 'purple'], [0.5, 'lightseagreen'], [1, 'gold']],
showscale=True,
cmin=-5.5,
cmax=3),
dimensions=list([
dict(range=[0, 3],
constraintrange=[0, 3],
label='Cluster',
values=sample['K_Cluster']),
dict(range=[-5.5, 1.5],
constraintrange=[-5.5, 1.5],
label='Recency',
values=sample['Recency']),
dict(range=[-.5, 2.5],
constraintrange=[-.5, 2.5],
label='Frequency',
values=sample['Frequency']),
dict(range=[-3, 3],
constraintrange=[-3, 3],
label='Monetary',
values=sample['Monetary'])
])))
fig.update_layout(plot_bgcolor='white', paper_bgcolor='white')
return fig
def parallelPlot():
"""Plot parallel graph for selected variants"""
data = [
go.Parcoords(
line=dict(
color=WWR["WWR"],
colorscale='RdBu',
showscale=False, #True to show colorscale legend
reversescale=False,
),
dimensions=list([
dict(range=[0, 5],
tickvals=[1, 2, 3, 4],
label="Orientation",
values=plotOri,
ticktext=['East', 'North', 'South', 'West']),
assignDict(WWR),
assignDict(ShadingActive),
assignDict(TotalRadWindow),
assignDict(TotalRadEnteredWindow),
assignDict(TotalIntGain),
assignDict(TotalHeating),
assignDict(TotalCooling),
assignDict(TotalEnergy),
assignDict(DF),
assignDict(sDA)
]))
]
layout = go.Layout(
width=1880,
height=888) #plot_bgcolor = '#E5E5E5', paper_bgcolor = '#E5E5E5'
fig = go.Figure(data=data, layout=layout)
py.offline.plot(fig, filename="parallel.html")
def setUp(self):
# Construct initial scatter object
self.parcoords = go.Parcoords(name="parcoords A")
# Assert initial state
d1, d2 = strip_dict_params(
self.parcoords, {"type": "parcoords", "name": "parcoords A"}
)
assert d1 == d2
# Construct expected results
self.expected_toplevel = {
"type": "parcoords",
"name": "parcoords A",
"dimensions": [
{"values": [2, 3, 1], "visible": True},
{"values": [1, 2, 3], "label": "dim1"},
],
}
self.layout = go.Layout()
self.expected_layout1 = {"updatemenus": [{}, {"font": {"family": "courier"}}]}
self.expected_layout2 = {
"updatemenus": [{}, {"buttons": [{}, {}, {"method": "restyle"}]}]
}
def generate_plot(path, auto_open=True):
df = pd.read_csv('result/' + path + '.csv')
min_max = df['score'].agg(['min', 'max'])
dimensions = []
for dimension in df.columns.values:
if dimension != 'score':
dimensions.append(
dict(range=[0, 1], label=dimension, values=df[dimension]))
else:
dimensions.append(
dict(range=[0, min_max.max()],
label=dimension,
values=df[dimension]))
line = dict(color=df['score'],
colorscale='Viridis',
showscale=True,
reversescale=True,
cmin=0,
cmax=min_max.max())
data = [go.Parcoords(line=line, dimensions=dimensions)]
layout = go.Layout(plot_bgcolor='#E5E5E5', paper_bgcolor='#E5E5E5')
fig = go.Figure(data=data, layout=layout)
plotly.offline.plot(fig,
filename='result/' + path + '.html',
auto_open=auto_open)
def parallel_coordinate(dictionary):
color_list = []
for i in dictionary:
number_of_rows = len(dictionary[i])
for j in range(number_of_rows):
color_list.append((-1) * j)
break
dimensions = list()
for key in dictionary:
if key != 'commit_id':
temp = dict()
temp['range'] = [min(dictionary[key]), max(dictionary[key])]
temp['label'] = key
temp['values'] = dictionary[key]
temp['tickformat'] = '0.2f'
dimensions.append(temp)
data3 = [
go.Parcoords(
line=dict(
color=color_list,
colorscale='Jet',
showscale=True,
reversescale=True,
cmin=-200, # value =200k
cmax=0),
dimensions=dimensions)
]
fig3 = go.Figure(data=data3)
return fig3
def drawParallelCoords(data, accuracyName, dimensionNames, lowerisbetter=True):
accuracies = [d[accuracyName] for d in data]
accuracy_dim = go.parcats.Dimension(
values=accuracies,
label="mae",
)
if lowerisbetter:
cmin = np.min(accuracies)
cmax = np.mean(accuracies) + np.std(accuracies)
else:
cmin = np.mean(accuracies) - np.std(accuracies)
cmax = np.max(accuracies)
cs = [int(acc / np.max(accuracies) * 255.) for acc in accuracies]
colors = ['rgb({},0,0)'.format(c) for c in cs]
dimensionList = []
for dimensionName in dimensionNames:
d = [d[dimensionName] for d in data]
dimensionList.append(dict(range=[np.min(d), np.max(d)], label=dimensionName,values=d))
dimensionList.append(dict(range=[np.min(accuracies), np.max(accuracies)],label=accuracyName,values=accuracies))
data = [go.Parcoords(
line = dict(color = accuracies,
colorscale = 'Bluered',
reversescale=lowerisbetter,
showscale = True,
cmin = cmin,
cmax = cmax),
dimensions = list(dimensionList))]
return data
def createFigureWidget(self):
dimensions = self.dimensions
data_lines = []
i = 0
if self.only_subsets == False:
colors = [i for r in range(self.data.size)]
colorscale = [[0, '#EEEEEE']]
else:
colors = []
colorscale = []
i = -1
for sset in self.data.subsets:
i = i + 1
tmplist = [i for r in range(len(sset.to_index_list()))]
colors.extend(tmplist)
colorscale.append([i, sset.style.color])
t_color = len(colorscale)
if (t_color > 1):
for c in colorscale:
c[0] = c[0] / (t_color - 1)
else:
colorscale = [[0, '#EEEEEE'], [1, '#EEEEEE']]
for dimension in dimensions:
line = {}
if hasattr(self.data[dimension].flatten(), 'codes'):
line['values'] = self.data[dimension].flatten().codes.tolist()
tickvals, tickmask = np.unique(
self.data[dimension].flatten().codes, return_index=True)
ticktext = self.data[dimension][tickmask]
line['tickvals'] = tickvals.tolist()
line['ticktext'] = ticktext.tolist()
else:
line['values'] = self.data[dimension].flatten().tolist()
line['label'] = dimension
for sset in self.data.subsets:
if hasattr(sset[dimension].flatten(), 'codes'):
tmplist = sset[dimension].codes.tolist()
else:
tmplist = sset[dimension].tolist()
line['values'].extend(tmplist)
data_lines.append(line)
data = [
go.Parcoords(line=dict(color=colors, colorscale=colorscale),
dimensions=data_lines),
]
layout = {
'title': self.options['title'].value,
'xaxis': {
'title': self.options['xaxis'].value,
},
'yaxis': {
'title': self.options['yaxis'].value,
}
}
return FigureWidget(data=data, layout=layout)
def add_trace(self, df):
self.traces.append(go.Parcoords(
line = dict(
color=df['target'],
colorscale='Jet',
),
dimensions = list(
[dict(label = col, values = df[col].values) for col in df.columns]
)))
def draw_parallel_coordinate(self):
"""Draw the parallel coordinate graph."""
# Graph the processed data.
data = self.all_data
# Create the dimensions for parallel coordinates.
dimensions = [
dict(
label=column,
values=data[column],
range=[
math.floor(min(data[column]) * 0.95),
math.ceil(max(data[column]) * 1.05)
]
)
for column in data.columns[:-1]
]
# Find the number of toads selected.
toad_num = len(self.selected_toad)
# Pick proper color.
color = colorlover.scales["7"]["qual"]["Dark2"]
# Find the color scale list.
color_scale = [
[1 * index / (toad_num - 1), color[index]]
for index in range(toad_num)
] if toad_num > 1 else [[0, color[0]]]
# Create the parallel coordinate.
parallel_coordinate = [
go.Parcoords(
line=dict(
color=data["Numerical ID"],
colorscale=color_scale
),
dimensions=dimensions
)
]
layout = go.Layout(height=750)
# Get the color toad matches.
color_toad = [
[color[index], toad]
for index, toad in enumerate(self.selected_toad)
]
# Return the html div to frontend.
return jsonify(
plot=plot(
go.Figure(data=parallel_coordinate, layout=layout),
show_link=False,
output_type="div",
include_plotlyjs=False
),
color_toad=color_toad
)
def __generate_trace(self,
points: pd.DataFrame,
legend: str,
metadata: list = None,
normalize: bool = False,
**kwargs):
dimension = points.shape[1]
# tweak points size for 3D plots
marker_size = 8
if dimension == 3:
marker_size = 4
# if indicated, perform normalization
if normalize:
points = (points - points.min()) / (points.max() - points.min())
marker = dict(color="#236FA4",
size=marker_size,
symbol="circle",
line=dict(color="#236FA4", width=1),
opacity=0.8)
marker.update(**kwargs)
if dimension == 2:
trace = go.Scattergl(x=points[0],
y=points[1],
mode="markers",
marker=marker,
name=legend,
customdata=metadata)
elif dimension == 3:
trace = go.Scatter3d(x=points[0],
y=points[1],
z=points[2],
mode="markers",
marker=marker,
name=legend,
customdata=metadata)
else:
dimensions = list()
for column in points:
dimensions.append(
dict(
range=[0, 1],
label=self.axis_labels[column:column + 1][0]
if self.axis_labels[column:column + 1] else None,
values=points[column],
))
trace = go.Parcoords(
line=dict(color="#236FA4"),
dimensions=dimensions,
name=legend,
)
return trace
def main():
dataTime, dataAccuracy = read()
dataTime = extract(dataTime, 'meanTime')
dataAccuracy = extract(dataAccuracy, 'corrects')
df = mkframe(dataAccuracy)
col = []
length = 18
cm = plt.get_cmap('gist_rainbow')
for i in range(length):
col.append(cm(1. * i / length)[:-1])
color = []
for i in col:
hexes = []
for j in i:
tmp = format(int(str(int(round(j * 255))), 10), 'x')
if len(tmp) != 2:
tmp = '0' + tmp
hexes.append(copy.deepcopy(tmp))
hex = '#' + hexes[0]
hex += hexes[1]
hex += hexes[2]
color.append([1.0, hex])
print((color))
print(df)
data = [
go.Parcoords(
line=dict(color=df['measure'], colorscale=color),
dimensions=list([
dict(
range=[-1, 1],
# name=df['The number of groups'],
label='Task 2',
values=df['task2_correlation']),
dict(
range=[-1, 1],
# name=df['The number of groups'],
label='Task 3',
values=df['task3_correlation']),
dict(
range=[-1, 1],
# name=df['The number of groups'],
label='Task 4',
values=df['task4_correlation'])
]))
]
layout = go.Layout(showlegend=True
# plot_bgcolor='#E5E5E5',
# paper_bgcolor='#E5E5E5'
)
# plt.figure()
# parallel_coordinates(data, 'The number of groups')
# plt.show()
fig = go.Figure(data=data, layout=layout)
plotly.offline.plot(fig)
def setUp(self):
# Construct with mocked _send_restyle_msg method
self.figure = go.Figure(data=[
go.Scatter(),
go.Bar(),
go.Parcoords(dimensions=[{}, {'label': 'dim 2'}, {}])
])
# Mock out the message method
self.figure._send_restyle_msg = MagicMock()
def parallel_coordinate_plot(self, investigate=None, image_name="Plot"):
"""
Plotly parallel coordinate plot
:param investigate: parameter in the legend
:param image_name: name of the image
:return:
"""
try:
from plotly.offline import plot
from plotly import tools
import plotly.graph_objs as go
except ImportError:
print('plotly is not installed')
cols = list(self._df)
has_category = False
if investigate is not None and investigate in self._df.columns:
cols.remove(investigate)
has_category = True
plotly_list = list()
for col in cols:
data_dict = dict(range=[self._df[col].min(), self._df[col].max()],
label=col,
values=self._df[col],
tickformat='.2f')
plotly_list.append(data_dict)
data = list()
if has_category:
data.append(
go.Parcoords(line=dict(color=self._df[investigate],
colorscale='Viridis',
showscale=True,
colorbar=dict(title=investigate)),
dimensions=plotly_list))
else:
data.append(go.Parcoords(dimensions=plotly_list))
# fig = go.Figure(data=data)
dir = os.path.dirname(os.path.dirname(os.path.dirname(__file__)))
plot(data, filename=dir + '/' + image_name + '.html')
def main(infile, col_dimensions, categorized, col_color):
"""
Produce an interactive paracords plotting html
Args:
infile: str, tabular file
col_dimensions: str, comma separated index numbers. For example: "3,4,5"
col_color: str, index number
"""
df = pd.read_csv(infile, sep='\t', parse_dates=True)
dimensions = []
col_dimensions = [int(x) - 1 for x in col_dimensions.split(',')]
for col in col_dimensions:
values = df[df.columns[col]]
if categorized == 'boolfalse' and all(type(e) is int for e in values):
dimensions.append(
dict(values=values, tickformat=",.2r", label=df.columns[col]))
elif categorized == 'boolfalse' and all(
type(e) is float for e in values):
dimensions.append(
dict(values=values, tickformat="g", label=df.columns[col]))
else:
unique_values = list(set(values))
unique_values.sort()
dimensions.append(
dict(range=[0, len(unique_values) - 1],
tickvals=list(range(len(unique_values))),
ticktext=[str(e) for e in unique_values],
values=list(map(lambda e: unique_values.index(e),
values)),
label=df.columns[col]))
col_color = int(col_color) - 1
colors = df[df.columns[col_color]]
if all(type(e) is int for e in colors):
tickformat = ",.2r"
elif all(type(e) is float for e in colors):
tickformat = "g"
else:
sys.exit(
"Error: the column for coloring must contain all numerical values!"
)
dimensions.append(
dict(values=colors, tickformat=tickformat,
label=df.columns[col_color]))
line = dict(color=colors,
colorscale='Jet',
showscale=True,
reversescale=True)
data = [go.Parcoords(line=line, dimensions=dimensions)]
plotly.offline.plot(data, filename="output.html", auto_open=False)
def update_linesPlot(ats, selecteddata):
if target in numericals:
fig = {
'data': [
go.Parcoords(line=dict(color=df[target],
colorscale='Viridis',
showscale=True),
dimensions=list([
dict(range=[min(df[i]),
max(df[i])],
label=i,
values=df[i]) for i in ats
]))
]
}
if target not in numericals:
fig = {
'data': [
go.Parcoords(line=dict(
color=df["numerical_target"],
colorbar=dict(
title=target,
titleside='top',
tickmode='array',
tickvals=[numtar[i] for i in set(df[target])],
ticktext=list(set(df[target])),
ticks='outside'),
colorscale='Viridis',
showscale=True),
dimensions=list([
dict(range=[min(df[i]),
max(df[i])],
label=i,
values=df[i]) for i in ats
]))
]
}
return fig
def prepare_parcoords():
data = [
go.Parcoords(
line=dict(color=df['professor'], # color bar based on Professors
# colorscale = 'Cividis',
reversescale=True,
showscale=False,
),
dimensions=list([
dict(
label='professor', values=df['professor']),
dict(
label='lecture', values=df['lecture']),
dict(
label='participants', values=df['participants']),
dict(
label='professional_expertise', values=df['professional_expertise']),
dict(
label='motivation', values=df['motivation']),
dict(
label='clear_presentation', values=df['clear_presentation']),
dict(
label='overall_impression', values=df['overall_impression']),
])
)
]
layout = go.Layout(
title=go.layout.Title(
text='Professor Review Distribution',
),
font=dict(
family="Arial,Times New Roman,Balto,Courier New,Gravitas One,Old Standard TT",
size=16,
color='#000000'
),
# margin=go.layout.Margin(
# t=200,
# pad=4
# ),
autosize=False,
width=1320,
height=800
)
# return iplot(data = data,layout = layout)
fig = go.Figure(data = data,layout = layout)
return fig
def plot(self, class_column, colorscale='RdBu', labels=None, ranges=None):
"""
Build a parallel coordinates plot.
:param class_column: Column of the pandas dataframe expected to be the "class", i.e, for every
unique value on this column, one line color.
:param colorscale: plotly colorscale (https://plot.ly/python/colorscales/).
:param labels: Name of every dimension.
:param ranges: Range to be plotted in every dimension.
:return:
"""
if labels is not None:
if len(labels) != len(list(self.df)) - 1:
raise ValueError("'labels' must be a list of length equal to the number of columns"
" of the dataframe minus one.")
else:
labels = []
for column in list(self.df):
labels.append(column)
if ranges is not None:
if len(ranges) != len(list(self.df)) - 1:
raise ValueError("'ranges' must be a list of length equal to the number of columns"
" of the dataframe minus one.")
else:
ranges = []
for column in list(self.df):
ranges.append([min(self.df[column]), max(self.df[column])])
dimensions = []
for index, column in enumerate(list(self.df)):
if not column == class_column:
dimensions.append(dict(
range=ranges[index],
label=labels[index],
values=self.df[column]
))
colormap = {k: i for i, k in enumerate(self.df[class_column].unique())}
self.data = [
go.Parcoords(
line=dict(
color = list(map(lambda x: colormap[x], self.df[class_column])),
colorscale=colorscale,
showscale=True
),
dimensions=dimensions
)
]
self._fig = go.Figure(data=self.data, layout=self.layout)
return self.plotting_method(self._fig)
def __generate_trace(self,
objectives: DataFrame,
metadata: list = None,
legend: str = '',
normalize: bool = False,
**kwargs):
number_of_objectives = objectives.shape[1]
if normalize:
objectives = (objectives - objectives.min()) / (objectives.max() -
objectives.min())
marker = dict(color='rgb(127, 127, 127)',
size=3,
symbol='x',
line=dict(color='rgb(204, 204, 204)', width=1),
opacity=0.8)
marker.update(**kwargs)
if number_of_objectives == 2:
trace = go.Scattergl(x=objectives[0],
y=objectives[1],
mode='markers',
marker=marker,
name=legend,
customdata=metadata)
elif number_of_objectives == 3:
trace = go.Scatter3d(x=objectives[0],
y=objectives[1],
z=objectives[2],
mode='markers',
marker=marker,
name=legend,
customdata=metadata)
else:
dimensions = list()
for column in objectives:
dimensions.append(
dict(range=[0, 1],
label=self.axis_labels[column:column + 1][0]
if self.axis_labels[column:column + 1] else None,
values=objectives[column]))
trace = go.Parcoords(
line=dict(color='blue'),
dimensions=dimensions,
name=legend,
)
return trace
def generate_parallel_coordinates(selected_packages, selected_measures):
data_list = []
for package_name in selected_packages:
data = lib.get_processed_package(package_name=package_name)
data_list.append(data)
df = pd.DataFrame(data=data_list)
dimensions = []
for measure in selected_measures:
measure_value = df[measure]
measure_range = [max(measure_value), min(measure_value)] if measure == "dependency_count" or measure == "open_issues_count" else [min(measure_value), max(measure_value)]
dimensions.append(dict(range =measure_range,
label = find_measure_details_by_name(measure)['label'], values=measure_value, tickvals=np.unique(measure_value).tolist()))
palette = get_palette(df.shape[0])
parcoords = go.Parcoords(
dimensions = list(dimensions),
line = dict(color = df.index,
colorscale = palette),
tickfont = dict(size=18),
labelfont = dict(size=18)
)
data = [parcoords]
layout = go.Layout(
plot_bgcolor='#171b26',
paper_bgcolor='#171b26',
font_color='#FFF',
font_size=18,
font_family='"Open Sans", "HelveticaNeue", "Helvetica Neue", Helvetica, Arial, sans-serif',
xaxis=dict(
showticklabels=False
),
yaxis=dict(
showticklabels=False
)
)
for i, package in enumerate(selected_packages):
trace_dummy = go.Scatter(
x=[0, 0, 0], # Data is irrelevant since it won't be shown
y=[0, 0, 0],
name=package,
showlegend=True,
marker=dict(color=palette[i], size=30)
)
data.append(trace_dummy)
return {'data': data, 'layout': layout}
def _get_par_coords(self) -> go.Parcoords:
config = Config()
self.trace: go.Parcoords = go.Parcoords(
line=dict(
color=config.color_scale[1][1],
colorscale=config.color_scale,
cmin=0,
cmax=1,
),
dimensions=[
self._get_dimension_dict(col) for col in self.selected_columns
],
)
return self.trace
def parallel_coordinates_plot(x_dict, y_data, y_label, lower_is_better=True):
assert (isinstance(x_dict, dict)), 'x_dict should be a dictionary'
assert (isinstance(y_data, list)), 'y_data should be a list'
if lower_is_better:
cmin = np.min(y_data)
cmax = np.mean(y_data) + np.std(y_data)
else:
cmin = np.mean(y_data) - np.std(y_data)
cmax = np.max(y_data)
dimensionList = []
for key in x_dict:
if True in [isinstance(element, str) for element in x_dict[key]]:
x_dict[key] = [str(element) for element in x_dict[key]]
categories = list(set(x_dict[key]))
values = [categories.index(element) for element in x_dict[key]]
tickvals = [-0.5] + list(range(
0, len(categories))) + [len(categories) - 0.5]
categories = [''] + categories + ['']
dimensionList.append(
dict(range=[min(tickvals), max(tickvals)],
tickvals=tickvals,
ticktext=categories,
label=key,
values=values))
else:
dimensionList.append(
dict(range=[np.min(x_dict[key]),
np.max(x_dict[key])],
label=key,
values=x_dict[key]))
dimensionList.append(
dict(range=[np.min(y_data), np.max(y_data)],
label=y_label,
values=y_data))
f = [
go.Parcoords(line=dict(color=y_data,
colorscale='Bluered',
reversescale=lower_is_better,
showscale=True,
cmin=cmin,
cmax=cmax),
dimensions=list(dimensionList))
]
return f
def plot_trials(spark_trials):
data = [
get_result_dict(t) for t in spark_trials.trials
if t['result']['status'] == 'ok' and not np.isnan(t['result']['loss'])
]
df = pd.DataFrame(data=data)
df = add_color_col(df)
plotly_data = [
go.Parcoords(line=dict(color=df['color'],
colorscale='Jet',
showscale=True,
reversescale=True,
cmin=-4000,
cmax=-100),
dimensions=get_dimensions(df))
]
displayHTML(plot(plotly_data, output_type='div'))