def plot(self) -> List[dcc.Graph]:
apps = self.model.apps
data_compounds = [
dc
for dc in self.model.get_all_data_compounds_with_sane_data()
if dc.get_aggregation_level() == AggregationLevel.APP
]
self.print_latex_table(data_compounds)
html_elements = []
html_elements.append(header_metric("Distribution Table"))
dt = dash_table.DataTable(
data=self.get_table_data(data_compounds),
columns=self.get_columns(),
style_cell={
'textAlign': 'left',
},
style_header={
'backgroundColor': colorconstants.RGB_COLOR_LIGHT_GREY,
},
merge_duplicate_headers=True,
# style_data_conditional=style_data_conditional,
)
html_elements.append(dt)
html_elements.append(header_metric("Histogram Plots"))
histogram_plots = self.get_histogram_plots(data_compounds)
html_elements.append(histogram_plots)
return html_elements
def plot(self):
html_elements = []
apps = sorted(self.model.apps)
matrix = self.load_data(apps)
df = pd.DataFrame({
'domain': [app.domain for app in apps],
'package_name': [app.package_name for app in apps],
'app': apps
})
by_domain = df.groupby("domain")
for domain, frame in by_domain:
domain_apps = [app for app in frame["app"]]
use_cases = sorted([
use_case.name for use_case in
self.model.use_case_manager.get_use_cases_for_domain(domain)
])
html_elements.append(
header_metric(f"Use Case Overview: {domain.value}"))
dt = self.get_overview_table(domain_apps, use_cases, matrix)
html_elements.append(dt)
bar_plots, layout = self.get_verified_use_case_execution_stacked_bar_charts(
apps, matrix)
html_elements.append(
header_metric(
"Use Case Executions Overview Stacked Bar Chart: Distribution of apps among verified use case executions"
))
graph = dcc.Graph(
id=USE_CASE_EXECUTIONS_OVERVIEW_STACKED_BAR_CHART_GRAPH_ID,
config=dict(displayModeBar=False),
figure=dict(data=bar_plots, layout=layout))
html_elements.append(graph)
bar_plots, layout = self.get_number_computed_verified_grouped_bar_chart(
apps, matrix)
html_elements.append(
header_metric(
"Use Case Executions Computed Verified Overview Stacked Bar Chart"
))
graph = dcc.Graph(
id=
USE_CASE_EXECUTIONS_COMPUTED_VERIFIED_OVERVIEW_STACKED_BAR_CHART_GRAPH_ID,
config=dict(displayModeBar=False),
figure=dict(data=bar_plots, layout=layout))
html_elements.append(graph)
return dbc.Container(id=self.get_ui_id(), children=html_elements)
def ui_element(cls, show_header=True):
assert cls.description is not None
children = []
if show_header:
children.append(header_metric(cls.description))
children.append(dcc.Loading(id=cls.get_ui_loading_id(), children=[], type="circle"))
return dbc.Container(children)
def get_correlation_graph(self):
apps = sorted(self.model.apps)
matrix = self.load_data(apps)
use_cases_all = self.model.use_case_manager.get_use_cases()
number_of_atds = self.calculate_number_of_atds(use_cases_all)
number_of_words = self.calculate_number_of_words(use_cases_all)
number_of_verified_uce = self.calculate_number_of_verified_use_case_executions(
apps, matrix, use_cases_all)
number_of_computed_uce = self.calculate_number_of_computed_use_case_executions(
apps, matrix, use_cases_all)
df = pd.DataFrame({
"Number of ATDs": number_of_atds,
"Number of words in use case": number_of_words,
"Number of computed UCEs": number_of_computed_uce,
"Number of verified UCEs": number_of_verified_uce,
})
# method = "pearson"
methods = [
'pearson', # Gaussian or Gaussian-like distribution
'spearman',
'kendall',
]
height = 500
width = 600
html_elements = []
for method in methods:
title = f'Correlation Matrix ({method})'
html_elements.append(header_metric(title))
corr = statisticsutil.calc_correlation_matrix(df, method)
heat, layout = plotutil.get_correlations_heatmap(corr)
graph = dcc.Graph(id=f"correlation-graph-{method}",
style={
'height': f'{height}px',
'width': f'{width}px',
},
config=dict(displayModeBar=False),
figure=dict(data=[heat], layout=layout))
if configutil.MATRICS_FIGURE_DUMP_PLOTS:
plotutil.write_plot_to_file(
[heat],
layout,
f"{USE_CASE_EXECUTIONS_CORRELATION_HEATMAP_GRAPH_ID}-{method}",
width=width,
height=height)
html_elements.append(graph)
return html_elements
def plot(self) -> List[dcc.Graph]:
html_elements = []
apps = self.model.apps
data_compounds: List[DataCompound] = [
dc for dc in self.model.get_all_data_compounds_with_sane_data()
if dc.get_aggregation_level() == AggregationLevel.APP and dc.name
not in PlayStoreCorrelationsMetricAppAL.filtered_metrics
]
for aggr_func, aggr_func_name in PlayStoreCorrelationsMetricAppAL.aggregation_functions:
# All valid metrics data for a correlation analysis
metrics_data = {
f"{dc.name}": self.aggregate_data(aggr_func, dc.data)
for dc in data_compounds
if len(self.aggregate_data(aggr_func, dc.data)) == len(apps)
}
assert len(metrics_data), f"Metrics data was empty"
metrics_df = pd.DataFrame(metrics_data)
for method in PlayStoreCorrelationsMetricAppAL.methods:
title = f'Correlation Matrix ({method}) ({aggr_func_name})'
html_elements.append(header_metric(title))
corr = statisticsutil.calc_correlation_matrix(
metrics_df, method, white_list)
heat, layout = plotutil.get_correlations_heatmap(
corr, triangle=False)
graph = dcc.Graph(
id=
f"play-store-correlation-graph-{method}-{aggr_func_name}",
style={
'height':
f'{PlayStoreCorrelationsMetricAppAL.height}px',
'width': f'{PlayStoreCorrelationsMetricAppAL.width}px',
},
config=dict(displayModeBar=False),
figure=dict(data=[heat], layout=layout))
if configutil.MATRICS_FIGURE_DUMP_PLOTS:
plotutil.write_plot_to_file(
[heat],
layout,
f"{PlayStoreCorrelationsMetricAppAL.get_id()}-{method}-{aggr_func_name}",
width=PlayStoreCorrelationsMetricAppAL.width,
height=PlayStoreCorrelationsMetricAppAL.height)
html_elements.append(graph)
return html_elements
def model_use_case_base_unmodified_graph_ctr(app_selectionv):
model_html_children = []
if hasattr(app, 'matrics_'):
model_metrics: ModelGraphs = app.matrics_.model.model_accessors[
modelmetrics.ModelGraphs.get_id()]
model_graphs = model_metrics.plot_use_case_base_unmodified_graph()
sel = app_selectionv
model_graph = list(
model_graphs.values())[0] if sel is None else model_graphs[sel]
model_html_children = [
header_metric("Base unmodified graph for use case identification"),
dbc.Container(id=f'Base use case transformed graph {sel}',
children=model_graph,
style={"border": "1px solid black"})
]
return model_html_children
def model_app_home_state_graph_ctr(app_selectionv):
model_html_children = []
if hasattr(app, 'matrics_'):
model_metrics: ModelGraphs = app.matrics_.model.model_accessors[
modelmetrics.ModelGraphs.get_id()]
model_graphs = model_metrics.get_app_home_state_graphs()
sel = app_selectionv
model_graph = list(
model_graphs.values())[0] if sel is None else model_graphs[sel]
model_html_children = [
header_metric(
"Base graph for use case identification with tagged app home state"
),
dbc.Container(id=f'model-app-home-state-graph-{sel}-div',
children=model_graph,
style={
"border": "1px solid black",
"width": configutil.MATRICS_VISUALIZATION_WIDTH
})
]
return model_html_children