def box(self, x, y):
if x and y:
return hv.BoxWhisker(self.data, kdims=[x],
vdims=[y]).opts(plot=self._plot_opts)
index = self.data.index.name or 'index'
df = self.data.reset_index()
df = pd.melt(df, id_vars=[index], var_name='Group', value_name='Value')
plot_opts = dict(self._plot_opts)
plot_opts['invert_axes'] = not self.kwds.get('vert', True)
return hv.BoxWhisker(df, kdims=['Group'],
vdims=['Value']).opts(plot=plot_opts)
def plot_boxwhiskers(df):
title = "Áramtermelés eloszlása zónánként"
boxwhisker = hv.BoxWhisker(df, ('ZONEID', 'Zóna'),
('POWER', 'Teljesítmény'),
label=title)
boxwhisker.options(show_legend=False, width=500, height=500)
title2 = "Áramtermelés eloszlása óránként csoportosítva"
boxwhisker2 = hv.BoxWhisker(df, ('HOUR', 'Óra'), ('POWER', 'Teljesítmény'),
label=title2)
boxwhisker2.options(show_legend=False, width=500, height=500)
options = {'BoxWhisker': dict(width=500, height=500)}
return (boxwhisker + boxwhisker2).options(options)
def boxInts(data):
p1hv = hv.BoxWhisker(data, ['clsLabel'],
'meanInt',
label="mean intensity per cluster").sort()
p1hv.opts(show_legend=False, width=500, cmap='Set1', ylim=(-1, 40))
# p1hv.opts(show_legend=False, width=500)
return p1hv
def boxplot_metrics(data,
title,
color_by='Method',
palette='Set3',
invert_xaxis=True):
data = data.melt(var_name='Variables', value_name='Metric Value')
data['Method'] = data['Variables'].apply(lambda x: x.split('-')[0])
data['Ranking'] = data['Variables'].apply(lambda x: x.split('-')[1])
data['Metric'] = data['Variables'].apply(lambda x: x.split('-')[2])
boxwhisker = hv.BoxWhisker(data, ['Metric', 'Ranking', 'Method'],
'Metric Value',
label=title)
boxwhisker.opts(show_legend=False,
width=900,
height=500,
box_fill_color=color_by,
cmap=palette,
ylim=(0, 1),
xrotation=45,
invert_xaxis=invert_xaxis,
toolbar='above')
f = hv.render(boxwhisker, backend='bokeh')
f.toolbar.logo = None
f.grid.grid_line_color = 'lightgrey'
f.grid.grid_line_dash = [5, 3]
return f
def create_cluster_boxplot(x_corpus: VectorizedCorpus,
token_clusters: pd.DataFrame, n_cluster: int,
color: str) -> hv.opts:
xs = np.arange(x_corpus.document_index.year.min(),
x_corpus.document_index.year.max() + 1, 1)
token_ids = list(
token_clusters[token_clusters.cluster == n_cluster].index)
Y = x_corpus.data[:, token_ids]
xsr = np.repeat(xs, Y.shape[1])
ysr = Y.ravel()
data = pd.DataFrame(data={'year': xsr, 'frequency': ysr})
violin: hv.BoxWhisker = hv.BoxWhisker(data, ('year', 'Year'),
('frequency', 'Frequency'))
violin_opts = {
'height': 600,
'width': 900,
'box_fill_color': color,
'xrotation': 90,
# 'violin_width': 0.8
}
return violin.opts(**violin_opts)
def BoxWhisker(dfs, measurement, configs, analysisType, **addConfigs):
"""Plots a measurement from all df as boxwisker"""
newConfigs = addConfigs
log.info("Generating BoxWhisker Plot for {}".format(measurement))
try:
plot = hv.BoxWhisker(
dfs["All"],
kdims="Name",
vdims=measurement,
label="BoxWhisker: {}".format(measurement),
group="BoxWhisker: {}".format(measurement),
)
# plot = relabelPlot(plot, label="{}".format(measurement))
# get labels from the configs
# ylabel = "{} [{}]".format(measurement, dfs[dfs["keys"][0]]["units"][dfs[dfs["keys"][0]]["measurements"].index(measurement)])
try:
ylabel = "{} [{}]".format(
measurement,
dfs[dfs["keys"][0]]["units"][dfs[
dfs["keys"][0]]["measurements"].index(measurement)],
)
except Exception as err:
log.error(
"Label could not be genereated for concatonated Histogram {}. Error: {}"
.format(measurement, err))
ylabel = "X-Axis"
plot.opts(
box_alpha=0.3,
xrotation=80,
box_color="blue",
height=500,
show_legend=False,
width=600,
whisker_color="blue",
ylabel=ylabel,
)
# Update the plot specific options if need be
generalOptions = configs[analysisType].get("General", {})
newConfigs.update(generalOptions.copy())
data_options = (configs[analysisType].get(measurement, {}).get(
"BoxWhisker", {}).get("PlotOptions", {}))
newConfigs.update(configs[analysisType].get(
"{}Options".format("BoxWhisker"), {}))
newConfigs.update(data_options)
plot = customize_plot(plot, "", configs[analysisType], **newConfigs)
except Exception as err:
log.error(
"Unexpected error happened during BoxWhisker plot generation {}. Error: {}"
.format(measurement, err))
return None
return plot
def hv_box(data,
label='',
style=dict(),
side='Successes',
width=800,
height=400):
return hv.BoxWhisker(data, ["Team", "Task"], side,
label=label).opts(plot=dict(tools=['hover'],
legend_position="top",
xrotation=45,
width=width,
height=height),
style=style)
def plot_boxplot_bokeh(df):
foo = df.stack().reset_index()
foo = foo[[foo.columns[1], 0]]
foo.columns = ['group', 'error']
table = hv.Table(foo)
hvplot = hv.BoxWhisker(table.data, kdims=['group'])
rend = hv.renderer('bokeh')
fig = rend.get_plot(hvplot).state
fig.plot_width = 600
fig.plot_height = 500
fig.xaxis.major_label_orientation = 45
fig.grid.grid_line_alpha = 0.5
fig.background_fill_color = 'beige'
fig.background_fill_alpha = 0.2
return fig
def boxplot_times(time_results):
aux = ['1', '0.9', '0.8', '0.7', '0.6', '0.5', '0.4', '0.3', '0.2', '0.1']
aux.reverse()
time_results.columns = aux
medias = np.mean(time_results)
prueba = time_results.melt()
box_times = hv.BoxWhisker(
prueba,
'variable',
'value',
label='Execution time for different number of edges')
times_plot = hv.Curve((aux, medias)).opts(color='red', width=400)
box_times = box_times.redim.label(variable='Percentage of the total edges',
value='Time (s)')
times_plot = times_plot.redim.label(x='Percentaje of the total edges')
return box_times * times_plot
def NeoRecoScatter(self):
# # A vs R colored by RecoPo
# p = figure(height=400)
# p.circle(x=x,y=y, size=3, color="navy", alpha=0.5)
out = self.recopoData.loc[
self.recopoData.NeoantigenRecognitionPotential != 0.0].copy()
# out.NeoantigenRecognitionPotential = np.log10(out.NeoantigenRecognitionPotential)
p = hv.BoxWhisker(out, 'Sample', 'NeoantigenRecognitionPotential')
p.opts(logy=True,
xrotation=90,
height=200,
box_fill_color='Sample',
show_legend=False,
cmap='Category20')
# p.opts(style=dict(box_color=hv.Cycle('Set1')))
o = hv.render(p)
return (o)
def cpu_box(data):
return hv.BoxWhisker(data, 'CPU', 'Utilization').relabel('CPU Usage')
])
TOOLS_rhat = [
"save", "pan", "reset", hover_tool_rhat, "ybox_zoom", "ywheel_zoom"
]
TOOLS_ess = [
"save", "pan", "reset", hover_tool_ess, "ybox_zoom", "ywheel_zoom"
]
TOOLS_mcse = ["save", "pan", "reset", hover_tool, "ybox_zoom", "ywheel_zoom"]
title = f"rhat accuracy ({datetime.now().date().isoformat()})"
boxwhisker = hv.BoxWhisker(
comparison[comparison.group == "rhat"].rename(columns={
"diff_min": "diff_min_rhat",
"diff": "diff_rhat"
}),
["diagnostic", "diff_min_rhat", "diff_max"],
["diff_rhat"],
label=title,
)
boxwhisker.opts(
show_legend=False,
width=400,
show_grid=True,
tools=TOOLS_rhat,
default_tools=[],
xlabel="",
)
title = f"ess ({datetime.now().date().isoformat()})"
boxwhisker2 = hv.BoxWhisker(
#*These can easily be created using numpy's histogram function as illustrated below. #The plot is interactive and comes with a bunch of tools. You can customize these tools as well; #for your many options, see http://bokeh.pydata.org/en/latest/docs/user_guide/tools.html. #example from bokeh.io import show import holoviews as hv import numpy as np frequencies, edges = np.histogram(flora['petal width'], bins = 5) hv.Histogram(frequencies, edges, label = 'Histogram') #2. ScatterPlot hv.Scatter(flora[['petal width','sepal length']],label = 'Scatter plot') #3. BoxPlot hv.BoxWhisker(flora['sepal length'], label = "Box whiskers plot") #Mid-level charts: the Plotting interface #Beyond the canned methods above, Bokeh provides a "mid-level" interface that more directly exposes the grammar of graphics methodology for constructing visual displays. #The basic procedure is #Create a blank canvas by calling bokeh.plotting.figure #Add glyphs, which are geometric shapes. #example from bokeh.plotting import figure # Create a canvas with a specific set of tools for the user: TOOLS = 'pan,box_zoom,wheel_zoom,lasso_select,save,reset,help'
def _fcn(key):
info = data[key]
if info.resolution.mean() < 1e-1:
info = info.assign(resolution=data[key].resolution * 1e3)
box = hv.BoxWhisker(info, ['trackcount', 'bead'], 'resolution')
return box.options(finalize_hooks=[_fhook])
def datafunc_down(data):
algo1_data = data['algo-1']
algo2_data = data['algo-2']
return hv.BoxWhisker(algo1_data, vdims='algo-1') + hv.BoxWhisker(
algo2_data, vdims='algo-2')
