Merge pull request #161 from kurusugawa-computer/modify-statistics-vi…

…sualize Modify statistics visualize
kurusugawa-computer · Jan 22, 2020 · 7509ed8 · 7509ed8
2 parents edf665b + 9650082
commit 7509ed8
Show file tree

Hide file tree

Showing 10 changed files with 294 additions and 212 deletions.
diff --git a/Pipfile.lock b/Pipfile.lock
diff --git a/annofabcli/__version__.py b/annofabcli/__version__.py
@@ -1 +1 @@
-__version__ = "1.18.1"
+__version__ = "1.18.2"
diff --git a/annofabcli/py.typed b/annofabcli/py.typed
diff --git a/annofabcli/statistics/graph.py b/annofabcli/statistics/graph.py
@@ -157,6 +157,9 @@ def write_histogram_for_worktime(self, df: pd.DataFrame):
 
         renderer = hv.renderer("bokeh")
 
+        output_file = f"{self.outdir}/html/{self.short_project_id}-ヒストグラム-作業時間"
+        logger.debug(f"{output_file}.html を出力します。")
+
         histogram_name_list = [
             HistogramName(column="annotation_worktime_hour", x_axis_label="教師付時間[hour]", title="教師付時間",),
             HistogramName(column="inspection_worktime_hour", x_axis_label="検査時間[hour]", title="検査時間",),
@@ -195,7 +198,7 @@ def write_histogram_for_worktime(self, df: pd.DataFrame):
 
         # 軸範囲が同期しないようにする
         layout = hv.Layout(histograms).options(shared_axes=False).cols(3)
-        renderer.save(layout, f"{self.outdir}/html/{self.short_project_id}-ヒストグラム-作業時間")
+        renderer.save(layout, output_file)
 
     def write_histogram_for_other(self, df: pd.DataFrame):
         """
@@ -211,6 +214,9 @@ def write_histogram_for_other(self, df: pd.DataFrame):
 
         renderer = hv.renderer("bokeh")
 
+        output_file = f"{self.outdir}/html/{self.short_project_id}-ヒストグラム"
+        logger.debug(f"{output_file}.html を出力します。")
+
         histogram_name_list = [
             HistogramName(column="annotation_count", x_axis_label="アノテーション数", title="アノテーション数"),
             HistogramName(column="input_data_count", x_axis_label="画像枚数", title="画像枚数"),
@@ -241,7 +247,7 @@ def write_histogram_for_other(self, df: pd.DataFrame):
 
         # 軸範囲が同期しないようにする
         layout = hv.Layout(histograms).options(shared_axes=False).cols(3)
-        renderer.save(layout, f"{self.outdir}/html/{self.short_project_id}-ヒストグラム")
+        renderer.save(layout, output_file)
 
     def write_histogram_for_annotation_count_by_label(self, df: pd.DataFrame) -> None:
         """
@@ -253,6 +259,9 @@ def write_histogram_for_annotation_count_by_label(self, df: pd.DataFrame) -> Non
 
         renderer = hv.renderer("bokeh")
 
+        output_file = f"{self.outdir}/html/{self.short_project_id}-ヒストグラム-ラベルごとのアノテーション数"
+        logger.debug(f"{output_file}.html を出力します。")
+
         histograms = []
         label_columns = [e for e in df.columns if e.startswith("label_")]
 
@@ -267,7 +276,7 @@ def write_histogram_for_annotation_count_by_label(self, df: pd.DataFrame) -> Non
 
         # 軸範囲が同期しないようにする
         layout = hv.Layout(histograms).options(shared_axes=False).cols(3)
-        renderer.save(layout, f"{self.outdir}/html/{self.short_project_id}-ヒストグラム-ラベルごとのアノテーション数")
+        renderer.save(layout, output_file)
 
     def create_user_id_list(
         self, df: pd.DataFrame, user_id_column: str, arg_user_id_list: Optional[List[str]] = None,
@@ -321,6 +330,9 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             Returns:
 
             """
+            output_file = f"{self.outdir}/html/{self.short_project_id}-{html_title}.html"
+            logger.debug(f"{output_file} を出力します。")
+
             figs: List[bokeh.plotting.Figure] = []
             for fig_info in fig_info_list:
                 figs.append(
@@ -354,9 +366,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
                 self._set_legend(fig, hover_tool)
 
             bokeh.plotting.reset_output()
-            bokeh.plotting.output_file(
-                f"{self.outdir}/html/{self.short_project_id}-{html_title}.html", title=html_title,
-            )
+            bokeh.plotting.output_file(output_file, title=html_title)
             bokeh.plotting.save(bokeh.layouts.column(figs))
 
         tooltip_item = [
@@ -380,7 +390,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
         first_annotation_user_id_list = self.create_user_id_list(
             df, "first_annotation_user_id", first_annotation_user_id_list
         )
-        logger.debug(f"グラフに表示するuser_id = {first_annotation_user_id_list}")
+        logger.debug(f"教師付者用の折れ線グラフに表示する、教師付者のuser_id = {first_annotation_user_id_list}")
 
         df["date_first_annotation_started_date"] = df["first_annotation_started_date"].map(
             lambda e: dateutil.parser.parse(e).date()
@@ -530,6 +540,9 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             Returns:
 
             """
+            output_file = f"{self.outdir}/html/{self.short_project_id}-{html_title}.html"
+            logger.debug(f"{output_file} を出力します。")
+
             figs: List[bokeh.plotting.Figure] = []
             for fig_info in fig_info_list:
                 figs.append(
@@ -577,9 +590,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
                 self._set_legend(fig, hover_tool)
 
             bokeh.plotting.reset_output()
-            bokeh.plotting.output_file(
-                f"{self.outdir}/html/{self.short_project_id}-{html_title}.html", title=html_title,
-            )
+            bokeh.plotting.output_file(output_file, title=html_title)
             bokeh.plotting.save(bokeh.layouts.column(figs))
 
         tooltip_item = [
@@ -605,7 +616,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
         first_annotation_user_id_list = self.create_user_id_list(
             df, "first_annotation_user_id", first_annotation_user_id_list
         )
-        logger.debug(f"グラフに表示するuser_id = {first_annotation_user_id_list}")
+        logger.debug(f"教師付者用の累計折れ線グラフに表示する、教師付者のuser_id = {first_annotation_user_id_list}")
 
         # 横軸が累計のアノテーション数
         fig_info_list_annotation_count = [
@@ -752,6 +763,9 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             Returns:
 
             """
+            output_file = f"{self.outdir}/html/{self.short_project_id}-{html_title}.html"
+            logger.debug(f"{output_file} を出力します。")
+
             figs: List[bokeh.plotting.Figure] = []
             for fig_info in fig_info_list:
                 figs.append(
@@ -796,9 +810,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
                 self._set_legend(fig, hover_tool)
 
             bokeh.plotting.reset_output()
-            bokeh.plotting.output_file(
-                f"{self.outdir}/html/{self.short_project_id}-{html_title}.html", title=html_title,
-            )
+            bokeh.plotting.output_file(output_file, title=html_title)
             bokeh.plotting.save(bokeh.layouts.column(figs))
 
         tooltip_item = [
@@ -830,7 +842,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             logger.info(f"検査フェーズを担当してユーザがいないため、検査者用のグラフは出力しません。")
             return
 
-        logger.debug(f"グラフに表示するuser_id = {first_inspection_user_id_list}")
+        logger.debug(f"検査者用の累計折れ線グラフに表示する、検査者のuser_id = {first_inspection_user_id_list}")
 
         # 横軸が累計のアノテーション数
         fig_info_list_annotation_count = [
@@ -909,6 +921,9 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             Returns:
 
             """
+            output_file = f"{self.outdir}/html/{self.short_project_id}-{html_title}.html"
+            logger.debug(f"{output_file} を出力します。")
+
             figs: List[bokeh.plotting.Figure] = []
             for fig_info in fig_info_list:
                 figs.append(
@@ -952,9 +967,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
                 self._set_legend(fig, hover_tool)
 
             bokeh.plotting.reset_output()
-            bokeh.plotting.output_file(
-                f"{self.outdir}/html/{self.short_project_id}-{html_title}.html", title=html_title,
-            )
+            bokeh.plotting.output_file(output_file, title=html_title)
             bokeh.plotting.save(bokeh.layouts.column(figs))
 
         tooltip_item = [
@@ -988,7 +1001,7 @@ def write_cumulative_graph(fig_info_list: List[Dict[str, str]], html_title: str)
             logger.info(f"受入フェーズを担当してユーザがいないため、受入者用のグラフは出力しません。")
             return
 
-        logger.debug(f"グラフに表示するuser_id = {first_acceptance_user_id_list}")
+        logger.debug(f"受入者用の累計折れ線グラフに表示する、受入者のuser_id = {first_acceptance_user_id_list}")
 
         # 横軸が累計のアノテーション数
         fig_info_list_annotation_count = [

diff --git a/annofabcli/statistics/table.py b/annofabcli/statistics/table.py
@@ -13,6 +13,7 @@
     WorktimeStatisticsItem,
 )
 from annofabapi.models import InputDataId, Inspection, InspectionStatus, Task, TaskHistory, TaskPhase, TaskStatus
+from annofabapi.utils import get_task_history_index_skipped_acceptance, get_task_history_index_skipped_inspection
 from more_itertools import first_true
 
 import annofabcli
@@ -139,6 +140,23 @@ def _create_annotation_summary(self, annotation_list: List[SimpleAnnotationDetai
 
         return annotation_summary
 
+    @staticmethod
+    def operator_is_changed_by_phase(task_history_list: List[TaskHistory], phase: TaskPhase) -> bool:
+        """
+        フェーズ内の作業者が途中で変わったかどうか
+
+        Args:
+            task_history_list: タスク履歴List
+            phase: フェーズ
+
+        Returns:
+            Trueならばフェーズ内の作業者が途中で変わった
+        """
+        account_id_list = [
+            e["account_id"] for e in task_history_list if TaskPhase(e["phase"]) == phase and e["account_id"] is not None
+        ]
+        return len(set(account_id_list)) >= 2
+
     @staticmethod
     def _inspection_condition(inspection_arg, exclude_reply: bool, only_error_corrected: bool):
         """
@@ -392,18 +410,6 @@ def diff_days(ended_key: str, started_key: str) -> Optional[float]:
             else:
                 return None
 
-        def acceptance_is_skipped(arg_task_histories: List[TaskHistory]) -> bool:
-            skipped_histories = [
-                e
-                for e in arg_task_histories
-                if (
-                    e["phase"] == TaskPhase.ACCEPTANCE.value
-                    and e["account_id"] is None
-                    and annofabcli.utils.isoduration_to_hour(e["accumulated_labor_time_milliseconds"]) == 0
-                )
-            ]
-            return len(skipped_histories) > 0
-
         annotation_histories = [e for e in task_histories if e["phase"] == TaskPhase.ANNOTATION.value]
         inspection_histories = [e for e in task_histories if e["phase"] == TaskPhase.INSPECTION.value]
         acceptance_histories = [e for e in task_histories if e["phase"] == TaskPhase.ACCEPTANCE.value]
@@ -481,8 +487,12 @@ def acceptance_is_skipped(arg_task_histories: List[TaskHistory]) -> bool:
 
         task["diff_days_to_task_completed"] = diff_days("task_completed_datetime", "first_annotation_started_datetime")
 
-        # 自動受入されたか否か
-        task["acceptance_is_skipped"] = acceptance_is_skipped(task_histories)
+        # 抜取検査/抜取受入で、検査/受入がスキップされたか否か
+        task["acceptance_is_skipped"] = len(get_task_history_index_skipped_acceptance(task_histories)) > 0
+        task["inspection_is_skipped"] = len(get_task_history_index_skipped_inspection(task_histories)) > 0
+        task["annotator_is_changed"] = self.operator_is_changed_by_phase(task_histories, TaskPhase.ANNOTATION)
+        task["inspector_is_changed"] = self.operator_is_changed_by_phase(task_histories, TaskPhase.INSPECTION)
+        task["acceptor_is_changed"] = self.operator_is_changed_by_phase(task_histories, TaskPhase.ACCEPTANCE)
 
         return task
 

diff --git a/annofabcli/statistics/tsv.py b/annofabcli/statistics/tsv.py
@@ -37,7 +37,7 @@ def _write_csv(self, filename: str, df: pd.DataFrame) -> None:
         """
         output_path = Path(f"{self.outdir}/{filename}")
         output_path.parent.mkdir(exist_ok=True, parents=True)
-        logger.debug(f"{str(output_path)} 書き込み")
+        logger.debug(f"{str(output_path)} を出力します。")
         df.to_csv(str(output_path), sep=",", encoding="utf_8_sig", index=False)
 
     @staticmethod
@@ -146,6 +146,12 @@ def write_task_list(self, df: pd.DataFrame, dropped_columns: Optional[List[str]]
             "annotation_count",
             "inspection_count",
             "input_data_count_of_inspection",
+            # タスクの状態
+            "annotator_is_changed",
+            "inspector_is_changed",
+            "acceptor_is_changed",
+            "inspection_is_skipped",
+            "acceptance_is_skipped",
         ]
         required_columns = self._create_required_columns(df, prior_columns, dropped_columns)
         self._write_csv(f"{self.short_project_id}-タスクlist.csv", df[required_columns])
@@ -182,6 +188,30 @@ def write_task_history_list(self, df: pd.DataFrame, dropped_columns: Optional[Li
         required_columns = self._create_required_columns(df, prior_columns, dropped_columns)
         self._write_csv(f"{self.short_project_id}-タスク履歴list.csv", df[required_columns])
 
+    def write_task_count(self, df: pd.DataFrame) -> None:
+        """
+        タスク数の集計結果をCSVで出力する。
+
+        Args:
+            df: タスクListのDataFrame
+
+        """
+        columns = [
+            "task_count",
+            "annotator_is_changed",
+            "inspector_is_changed",
+            "acceptor_is_changed",
+            "inspection_is_skipped",
+            "acceptance_is_skipped",
+        ]
+
+        sum_series = df[columns].sum()
+        sum_df = pd.DataFrame()
+        sum_df["column"] = sum_series.index
+        sum_df["count_if_true"] = sum_series.values
+
+        self._write_csv(f"{self.short_project_id}-タスク数の集計.csv", sum_df)
+
     def write_member_list(self, df: pd.DataFrame, dropped_columns: Optional[List[str]] = None):
         """
         プロジェクトメンバ一覧をTSVで出力する