def action_distrubtion(episode, ref_episode):
actions_per_sub = actions_model.get_action_per_sub(episode)
actions_per_sub.append(actions_model.get_action_per_sub(ref_episode)[0])
y_max = None
if len(actions_per_sub[0]["y"]) == 0:
figure_subs = go.Figure(
layout=layout_no_data("No Actions on subs for this Agent"))
else:
figure_subs = go.Figure(layout=layout_def, data=actions_per_sub)
y_max = max(map(max_or_zero, [trace.y
for trace in actions_per_sub])) + 1
actions_per_lines = actions_model.get_action_per_line(episode)
actions_per_lines.append(actions_model.get_action_per_line(ref_episode)[0])
if len(actions_per_lines[0]["y"]) == 0:
figure_lines = go.Figure(
layout=layout_no_data("No Actions on lines for this Agent"))
else:
figure_lines = go.Figure(layout=layout_def, data=actions_per_lines)
if y_max is None:
y_max = max(
map(max_or_zero, [trace.y for trace in actions_per_lines])) + 1
if max(map(max_or_zero, [trace.y
for trace in actions_per_lines])) > y_max:
y_max = max(
map(max_or_zero, [trace.y for trace in actions_per_lines])) + 1
actions_redisp = actions_model.get_action_redispatch(episode)
actions_redisp.append(actions_model.get_action_redispatch(ref_episode)[0])
if len(actions_redisp[0]["y"]) == 0:
figure_redisp = go.Figure(
layout=layout_no_data("No redispatching actions for this Agent"))
else:
figure_redisp = go.Figure(layout=layout_def, data=actions_redisp)
if y_max is None:
y_max = max(map(max_or_zero, [trace.y
for trace in actions_redisp])) + 1
if max(map(max_or_zero, [trace.y
for trace in actions_redisp])) > y_max:
y_max = max(map(max_or_zero, [trace.y
for trace in actions_redisp])) + 1
if y_max:
figure_subs.update_yaxes(range=[0, y_max])
figure_lines.update_yaxes(range=[0, y_max])
figure_redisp.update_yaxes(range=[0, y_max])
return ActionsDistribution(on_subs=figure_subs,
on_lines=figure_lines,
redisp=figure_redisp)
def test_multi_topo(self):
self.agent_name = "multiTopology-baseline"
self.scenario_name = "000"
self.episode_data = EpisodeData.from_disk(
os.path.join(self.agents_path, self.agent_name), self.scenario_name
)
self.episode_analytics = EpisodeAnalytics(
self.episode_data, self.scenario_name, self.agent_name
)
nb_actions = self.episode_analytics.action_data_table[
["action_line", "action_subs"]
].sum()
self.assertEqual(nb_actions.action_line, 25.0)
self.assertEqual(nb_actions.action_subs, 38.0)
action_per_line = get_action_per_line(self.episode_analytics)
action_per_sub = get_action_per_sub(self.episode_analytics)
# We need to sort the labels for which values are equal.
# Otherwise, the output is random.
self.assertListEqual(sorted(action_per_sub[0].x.tolist()), ["sub_3", "sub_5"])
self.assertListEqual(action_per_sub[0].y.tolist(), [19, 19])
self.assertListEqual(action_per_line[0].x.tolist(), ["3_6_15", "9_10_12"])
self.assertListEqual(action_per_line[0].y.tolist(), [13, 12])
self.assertListEqual(
self.episode_analytics.action_data_table.action_id[:5].tolist(),
[0, 1, 1, 2, 3],
)
self.assertListEqual(
self.episode_analytics.action_data_table.distance[:5].tolist(),
[1, 2, 2, 0, 3],
)
def update_agent_log_action_graphs(study_agent, figure_sub, figure_switch_line,
scenario):
new_episode = make_episode(study_agent, scenario)
figure_sub["data"] = actions_model.get_action_per_sub(new_episode)
figure_switch_line["data"] = actions_model.get_action_per_line(new_episode)
figure_sub["layout"].update(
actions_model.update_layout(
len(figure_sub["data"][0]["x"]) == 0,
"No Actions on subs for this Agent"))
figure_switch_line["layout"].update(
actions_model.update_layout(
len(figure_switch_line["data"][0]["x"]) == 0,
"No Actions on lines for this Agent"))
return figure_sub, figure_switch_line
def action_distrubtion(episode):
actions_per_sub = actions_model.get_action_per_sub(episode)
if len(actions_per_sub[0]["y"]) == 0:
figure_subs = go.Figure(
layout=layout_no_data("No Actions on subs for this Agent"))
else:
figure_subs = go.Figure(layout=layout_def, data=actions_per_sub)
actions_per_lines = actions_model.get_action_per_line(episode)
if len(actions_per_lines[0]["y"]) == 0:
figure_lines = go.Figure(
layout=layout_no_data("No Actions on lines for this Agent"))
else:
figure_lines = go.Figure(layout=layout_def, data=actions_per_lines)
return ActionsDistribution(on_subs=figure_subs, on_lines=figure_lines)
def test_action_repartition(self):
nb_actions = self.episode_analytics.action_data_table[
["action_line", "action_subs"]
].sum()
self.assertEqual(nb_actions.action_line, 0.0)
self.assertEqual(nb_actions.action_subs, 31.0)
action_per_line = get_action_per_line(self.episode_analytics)
action_per_sub = get_action_per_sub(self.episode_analytics)
self.assertListEqual(action_per_line[0].x.tolist(), [])
# We need to sort the labels for which values are equal.
# Otherwise, the output is random.
self.assertListEqual(
[
*action_per_sub[0].x.tolist()[:-2],
*sorted(action_per_sub[0].x.tolist()[-2:]),
],
["sub_4", "sub_1", "sub_3", "sub_8", "sub_9"],
)
self.assertListEqual(action_per_sub[0].y.tolist(), [13, 8, 4, 3, 3])
def test_action_repartition(self):
nb_actions = self.episode_analytics.action_data_table[
['action_line', 'action_subs']].sum()
self.assertEqual(nb_actions.action_line, 0.0)
self.assertEqual(nb_actions.action_subs, 177.0)
action_per_line = get_action_per_line(self.episode_analytics)
action_per_sub = get_action_per_sub(self.episode_analytics)
self.assertListEqual(
action_per_line[0].x.tolist(),
[]
)
# We need to sort the labels for which values are equal.
# Otherwise, the output is random.
self.assertListEqual(
[*action_per_sub[0].x.tolist()[:-2], *sorted(action_per_sub[0].x.tolist()[-2:])],
['sub_9', 'sub_8', 'sub_4', 'sub_12', 'sub_1', 'sub_3']
)
self.assertListEqual(
action_per_sub[0].y.tolist(),
[106, 45, 12, 10, 2, 2]
)
def update_agent_log_action_graphs(study_agent, ref_agent, figure_sub,
figure_switch_line, figure_redisp,
scenario):
new_episode = make_episode(study_agent, scenario)
ref_episode = make_episode(ref_agent, scenario)
y_max = None
figure_sub["data"] = actions_model.get_action_per_sub(new_episode)
if len(figure_sub["data"][0]["x"]) != 0:
figure_sub["data"].append(
actions_model.get_action_per_sub(ref_episode)[0])
y_max = max(
map(max_or_zero, [trace.y
for trace in figure_sub["data"]])) + 1
figure_switch_line["data"] = actions_model.get_action_per_line(
new_episode)
if len(figure_switch_line["data"][0]["x"]) != 0:
figure_switch_line["data"].append(
actions_model.get_action_per_line(ref_episode)[0])
if y_max is None:
y_max = (max(
map(
max_or_zero,
[trace.y for trace in figure_switch_line["data"]],
)) + 1)
if (max(
map(max_or_zero,
[trace.y
for trace in figure_switch_line["data"]])) > y_max):
y_max = (max(
map(
max_or_zero,
[trace.y for trace in figure_switch_line["data"]],
)) + 1)
figure_redisp["data"] = actions_model.get_action_redispatch(
new_episode)
if len(figure_redisp["data"][0]["x"]) != 0:
figure_redisp["data"].append(
actions_model.get_action_redispatch(ref_episode)[0])
if y_max is None:
y_max = (max(
map(max_or_zero,
[trace.y for trace in figure_redisp["data"]])) + 1)
if (max(
map(max_or_zero,
[trace.y
for trace in figure_redisp["data"]])) > y_max):
y_max = (max(
map(max_or_zero,
[trace.y for trace in figure_redisp["data"]])) + 1)
figure_sub["layout"].update(
actions_model.update_layout(
len(figure_sub["data"][0]["x"]) == 0,
"No Actions on subs for this Agent",
))
figure_switch_line["layout"].update(
actions_model.update_layout(
len(figure_switch_line["data"][0]["x"]) == 0,
"No Actions on lines for this Agent",
))
figure_redisp["layout"].update(
actions_model.update_layout(
len(figure_redisp["data"][0]["x"]) == 0,
"No redispatching actions for this Agent",
))
if y_max:
figure_sub["layout"]["yaxis"].update(range=[0, y_max])
figure_switch_line["layout"]["yaxis"].update(range=[0, y_max])
figure_redisp["layout"]["yaxis"].update(range=[0, y_max])
return figure_sub, figure_switch_line, figure_redisp
def action_distrubtion(episode):
figure_subs = go.Figure(layout=layout_def,
data=actions_model.get_action_per_sub(episode))
figure_lines = go.Figure(layout=layout_def,
data=actions_model.get_action_per_line(episode))
return ActionsDistribution(on_subs=figure_subs, on_lines=figure_lines)
