feat: response time timeseries (#322)

* diff_node_names Signed-off-by: yu-taro- <[email protected]> * 20230130 Signed-off-by: yu-taro- <[email protected]> * diff_node_names 20230205 * re Signed-off-by: yu-taro- <[email protected]> * class Diff Signed-off-by: yu-taro- <[email protected]> * 20230205 Signed-off-by: yu-taro- <[email protected]> * refactor(arcitecture), test(test_arcitecture) / Naming conventions were changed and mock tests were added and modified because of the points raised Signed-off-by: yu-taro- <[email protected]> * a Signed-off-by: yu-taro- <[email protected]> * conflict fix2 Signed-off-by: yu-taro- <[email protected]> * feat: make response time time series plot func Signed-off-by: taro-yu <[email protected]> * fix: conflict Signed-off-by: taro-yu <[email protected]> * fix: changed to use new record api Signed-off-by: taro-yu <[email protected]> * fix: changed to use new record api Signed-off-by: taro-yu <[email protected]> * fix: pytest errors Signed-off-by: taro-yu <[email protected]> * docs: add some description Signed-off-by: taro-yu <[email protected]> * docs: add some description Signed-off-by: taro-yu <[email protected]> * fix: argument error Signed-off-by: taro-yu <[email protected]> * feat: add hover info Signed-off-by: taro-yu <[email protected]> * docs: fix a comment Signed-off-by: taro-yu <[email protected]> * docs: added CARET_doc and fixed spell-ceheck-errors Signed-off-by: taro-yu <[email protected]> * docs: fixed spell-check-error Signed-off-by: taro-yu <[email protected]> * fix: mypy errors Signed-off-by: taro-yu <[email protected]> * fix: mypy errors Signed-off-by: taro-yu <[email protected]> * fix: mypy errors Signed-off-by: taro-yu <[email protected]> * fix: mypy errors and add some docs Signed-off-by: taro-yu <[email protected]> * fix: pytest errors Signed-off-by: taro-yu <[email protected]> * fix: mypy errors Signed-off-by: taro-yu <[email protected]> * docs remove some docs Signed-off-by: taro-yu <[email protected]> * feat: change hover info Signed-off-by: taro-yu <[email protected]> --------- Signed-off-by: yu-taro- <[email protected]> Signed-off-by: taro-yu <[email protected]> Co-authored-by: taro-yu <[email protected]>
tier4 · Sep 13, 2023 · 7efeaba · 7efeaba
1 parent b9ad974
commit 7efeaba
Show file tree

Hide file tree

Showing 9 changed files with 210 additions and 28 deletions.
diff --git a/src/caret_analyze/plot/metrics_base.py b/src/caret_analyze/plot/metrics_base.py
@@ -20,9 +20,9 @@
 import pandas as pd
 
 from ..record import ColumnValue, Range, RecordFactory, RecordsFactory, RecordsInterface
-from ..runtime import CallbackBase, Communication, Publisher, Subscription
+from ..runtime import CallbackBase, Communication, Path, Publisher, Subscription
 
-TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription)
+TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription) | Path
 
 
 class MetricsBase(metaclass=ABCMeta):
@@ -65,6 +65,8 @@ def _convert_timeseries_records_to_sim_time(
                     break
             provider = converter_cb._provider
             converter = provider.get_sim_time_converter(frame_min, frame_max)
+        elif isinstance(self._target_objects[0], Path):
+            pass
         else:
             provider = self._target_objects[0]._provider
             converter = provider.get_sim_time_converter(frame_min, frame_max)
@@ -122,5 +124,7 @@ def _add_top_level_column(
                            f'{target_object.subscribe_node_name}')
         elif isinstance(target_object, (Publisher, Subscription)):
             column_name = target_object.topic_name
+        elif isinstance(target_object, Path):
+            column_name = f'{target_object.column_names[0]} to {target_object.column_names[-1]}'
 
         return pd.concat([target_df], keys=[column_name], axis=1)
diff --git a/src/caret_analyze/plot/plot_facade.py b/src/caret_analyze/plot/plot_facade.py
@@ -29,7 +29,7 @@
 
 logger = getLogger(__name__)
 
-TimeSeriesTypes = CallbackBase | Communication | Publisher | Subscription
+TimeSeriesTypes = CallbackBase | Communication | Publisher | Subscription | Path
 CallbackSchedTypes = (Application | Executor | Path |
                       Node | CallbackGroup | Collection[CallbackGroup])
 
@@ -174,6 +174,33 @@ def create_latency_timeseries_plot(
         )
         return plot
 
+    def create_response_time_timeseries_plot(
+        *target_objects: Path,
+        case: str = 'best'
+    ) -> PlotBase:
+        """
+        Get response time timeseries plot instance.
+
+        Parameters
+        ----------
+        target_objects : Collection[Path]
+            Instances that are the sources of the plotting.
+            This also accepts multiple inputs by unpacking.
+        case: str, optional
+            Response time calculation method, best by default.
+            supported case: [best/worst/all].
+
+        Returns
+        -------
+        PlotBase
+
+        """
+        visualize_lib = VisualizeLibFactory.create_instance()
+        plot = TimeSeriesPlotFactory.create_instance(
+            parse_collection_or_unpack(target_objects), 'response_time', visualize_lib, case
+        )
+        return plot
+
     @staticmethod
     def create_response_time_histogram_plot(
         *paths: Path,

diff --git a/src/caret_analyze/plot/timeseries/response_time_timeseries.py b/src/caret_analyze/plot/timeseries/response_time_timeseries.py
@@ -0,0 +1,115 @@
+# Copyright 2021 Research Institute of Systems Planning, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from __future__ import annotations
+
+from collections.abc import Sequence
+
+import pandas as pd
+
+from ..metrics_base import MetricsBase
+from ...record import RecordsInterface, ResponseTime
+from ...runtime import Path
+
+
+class ResponseTimeTimeSeries(MetricsBase):
+    """Class that provides latency timeseries data."""
+
+    def __init__(
+        self,
+        target_objects: Sequence[Path],
+        case: str
+    ) -> None:
+        super().__init__(target_objects)
+        self._case = case
+
+    def to_dataframe(self, xaxis_type: str = 'system_time') -> pd.DataFrame:
+        """
+        Get Response time timeseries data for each object in pandas DataFrame format.
+
+        Parameters
+        ----------
+        xaxis_type : str
+            Type of time for timestamp.
+            "system_time", "index", or "sim_time" can be specified.
+            The default is "system_time".
+
+        Returns
+        -------
+        pd.DataFrame
+            Multi-column Response time DataFrame.
+
+        Notes
+        -----
+        xaxis_type "system_time" and "index" return the same DataFrame.
+
+        """
+        timeseries_records_list = self.to_timeseries_records_list(xaxis_type)
+        all_df = pd.DataFrame()
+        for to, response_records in zip(self._target_objects, timeseries_records_list):
+            response_df = response_records.to_dataframe()
+            response_df[response_df.columns[-1]] *= 10**(-6)
+            response_df.rename(
+                columns={
+                    response_df.columns[0]: 'path_start_timestamp [ns]',
+                    response_df.columns[1]: 'response_time [ms]',
+                },
+                inplace=True
+            )
+            # TODO: Multi-column DataFrame are difficult for users to handle,
+            #       so it should be a single-column DataFrame.
+            response_df = self._add_top_level_column(response_df, to)
+            all_df = pd.concat([all_df, response_df], axis=1)
+
+        return all_df.sort_index(level=0, axis=1, sort_remaining=False)
+
+    def to_timeseries_records_list(
+        self,
+        xaxis_type: str = 'system_time'
+    ) -> list[RecordsInterface]:
+        """
+        Get Response time records list of all target objects.
+
+        Parameters
+        ----------
+        xaxis_type : str
+            Type of time for timestamp.
+            "system_time", "index", or "sim_time" can be specified.
+            The default is "system_time".
+
+        Returns
+        -------
+        list[RecordsInterface]
+            Response time records list of all target objects.
+
+        """
+        timeseries_records_list: list[RecordsInterface] = [
+            _.to_records() for _ in self._target_objects
+        ]
+
+        if xaxis_type == 'sim_time':
+            timeseries_records_list = \
+                self._convert_timeseries_records_to_sim_time(timeseries_records_list)
+
+        response_timeseries_list: list[RecordsInterface] = []
+        for records in timeseries_records_list:
+            response = ResponseTime(records)
+            if self._case == 'best':
+                response_timeseries_list.append(response.to_best_case_records())
+            elif self._case == 'worst':
+                response_timeseries_list.append(response.to_worst_case_records())
+            elif self._case == 'all':
+                response_timeseries_list.append(response.to_all_records())
+
+        return response_timeseries_list
diff --git a/src/caret_analyze/plot/timeseries/timeseries_plot.py b/src/caret_analyze/plot/timeseries/timeseries_plot.py
@@ -22,9 +22,9 @@
 from ..plot_base import PlotBase
 from ..visualize_lib import VisualizeLibInterface
 from ...exceptions import UnsupportedTypeError
-from ...runtime import CallbackBase, Communication, Publisher, Subscription
+from ...runtime import CallbackBase, Communication, Path, Publisher, Subscription
 
-TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription)
+TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription) | Path
 
 
 class TimeSeriesPlot(PlotBase):
@@ -33,10 +33,12 @@ class TimeSeriesPlot(PlotBase):
     def __init__(
         self,
         metrics: MetricsBase,
-        visualize_lib: VisualizeLibInterface
+        visualize_lib: VisualizeLibInterface,
+        case: str = 'best'  # case is only used for response time timeseries.
     ) -> None:
         self._metrics = metrics
         self._visualize_lib = visualize_lib
+        self._case = case
 
     def to_dataframe(self, xaxis_type: str = 'system_time') -> pd.DataFrame:
         """
@@ -104,7 +106,8 @@ def figure(
             self._metrics,
             xaxis_type,
             ywheel_zoom,
-            full_legends
+            full_legends,
+            self._case
         )
 
     def _validate_xaxis_type(self, xaxis_type: str) -> None:

diff --git a/src/caret_analyze/plot/timeseries/timeseries_plot_factory.py b/src/caret_analyze/plot/timeseries/timeseries_plot_factory.py
@@ -19,14 +19,15 @@
 from .frequency_timeseries import FrequencyTimeSeries
 from .latency_timeseries import LatencyTimeSeries
 from .period_timeseries import PeriodTimeSeries
+from .response_time_timeseries import ResponseTimeTimeSeries
 from .timeseries_plot import TimeSeriesPlot
 from ..metrics_base import MetricsBase
 from ..visualize_lib import VisualizeLibInterface
 from ...common import type_check_decorator
 from ...exceptions import UnsupportedTypeError
-from ...runtime import CallbackBase, Communication, Publisher, Subscription
+from ...runtime import CallbackBase, Communication, Path, Publisher, Subscription
 
-TimeSeriesPlotTypes = CallbackBase | Communication | (Publisher | Subscription)
+TimeSeriesPlotTypes = CallbackBase | Communication | (Publisher | Subscription) | Path
 
 
 class TimeSeriesPlotFactory:
@@ -37,20 +38,23 @@ class TimeSeriesPlotFactory:
     def create_instance(
         target_objects: Sequence[TimeSeriesPlotTypes],
         metrics: str,
-        visualize_lib: VisualizeLibInterface
+        visualize_lib: VisualizeLibInterface,
+        case: str = 'best'  # case is only used for response time timeseries.
     ) -> TimeSeriesPlot:
         """
         Create an instance of TimeSeriesPlot.
 
         Parameters
         ----------
         target_objects : Sequence[TimeSeriesPlotTypes]
-            TimeSeriesPlotTypes = CallbackBase | Communication | (Publisher | Subscription)
+            TimeSeriesPlotTypes = CallbackBase | Communication | (Publisher | Subscription) | Path
         metrics : str
             Metrics for timeseries data.
-            supported metrics: [frequency/latency/period]
+            supported metrics: [frequency/latency/period/response_time]
         visualize_lib : VisualizeLibInterface
             Instance of VisualizeLibInterface used for visualization.
+        case : str
+            Parameter specifying best, worst or all. Use to create Response time timeseries graph.
 
         Returns
         -------
@@ -59,22 +63,28 @@ def create_instance(
         Raises
         ------
         UnsupportedTypeError
-            Argument metrics is not "frequency", "latency", or "period".
+            Argument metrics is not "frequency", "latency", "period", or "response_time".
 
         """
         metrics_: MetricsBase
+        PlotTypes = (CallbackBase, Communication, Publisher, Subscription)
         if metrics == 'frequency':
-            metrics_ = FrequencyTimeSeries(list(target_objects))
+            metrics_ = FrequencyTimeSeries([_ for _ in target_objects if isinstance(_, PlotTypes)])
             return TimeSeriesPlot(metrics_, visualize_lib)
         elif metrics == 'latency':
             # Ignore the mypy type check because type_check_decorator is applied.
             metrics_ = LatencyTimeSeries(list(target_objects))  # type: ignore
             return TimeSeriesPlot(metrics_, visualize_lib)
         elif metrics == 'period':
-            metrics_ = PeriodTimeSeries(list(target_objects))
+            metrics_ = PeriodTimeSeries([_ for _ in target_objects if isinstance(_, PlotTypes)])
             return TimeSeriesPlot(metrics_, visualize_lib)
+        elif metrics == 'response_time':
+            metrics_ = ResponseTimeTimeSeries(
+                [_ for _ in target_objects if isinstance(_, Path)], case
+            )
+            return TimeSeriesPlot(metrics_, visualize_lib, case)
         else:
             raise UnsupportedTypeError(
                 'Unsupported metrics specified. '
-                'Supported metrics: [frequency/latency/period]'
+                'Supported metrics: [frequency/latency/period/response_time]'
             )
diff --git a/src/caret_analyze/plot/visualize_lib/bokeh/bokeh.py b/src/caret_analyze/plot/visualize_lib/bokeh/bokeh.py
@@ -28,7 +28,7 @@
 from ...metrics_base import MetricsBase
 from ....runtime import CallbackBase, CallbackGroup, Communication, Path, Publisher, Subscription
 
-TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription)
+TimeSeriesTypes = CallbackBase | Communication | (Publisher | Subscription) | Path
 
 logger = getLogger(__name__)
 
@@ -131,7 +131,8 @@ def timeseries(
         metrics: MetricsBase,
         xaxis_type: str,
         ywheel_zoom: bool,
-        full_legends: bool
+        full_legends: bool,
+        case: str
     ) -> Figure:
         """
         Get a timeseries figure.
@@ -150,12 +151,15 @@ def timeseries(
         full_legends : bool
             If True, all legends are drawn
             even if the number of legends exceeds the threshold.
+        case : str
+            Parameter specifying best, worst or all. Use to create Response time timeseries graph.
+
 
         Returns
         -------
         bokeh.plotting.Figure
             Figure of timeseries.
 
         """
-        timeseries = BokehTimeSeries(metrics, xaxis_type, ywheel_zoom, full_legends)
+        timeseries = BokehTimeSeries(metrics, xaxis_type, ywheel_zoom, full_legends, case)
         return timeseries.create_figure()