Merge pull request #2650 from AllenInstitute/rc/2.15.0

Rc/2.15.0

allensdk/__init__.py

@@ -35,7 +35,7 @@
 #
 import logging
 
-__version__ = '2.14.1'
+__version__ = '2.15.0'
 
 
 try:

allensdk/brain_observatory/behavior/behavior_project_cache/project_apis/data_io/behavior_project_lims_api.py

@@ -21,6 +21,7 @@
     build_in_list_selector_query, build_where_clause)
 from allensdk.internal.brain_observatory.util.multi_session_utils import \
     get_session_metadata_multiprocessing
+from allensdk.brain_observatory.behavior.data_objects import BehaviorSessionId
 
 
 class BehaviorProjectLimsApi(BehaviorProjectBase):
@@ -236,6 +237,7 @@ def _get_behavior_summary_table(self) -> pd.DataFrame:
         query = f"""
             SELECT
                 bs.id AS behavior_session_id,
+                pr.code as project_code,
                 equipment.name as equipment_name,
                 bs.date_of_acquisition,
                 d.id as donor_id,
@@ -257,6 +259,7 @@ def _get_behavior_summary_table(self) -> pd.DataFrame:
                 {self._build_line_from_donor_query("driver")}
             ) driver on driver.donor_id = d.id
             LEFT OUTER JOIN equipment ON equipment.id = bs.equipment_id
+            LEFT OUTER JOIN projects pr ON pr.id = bs.project_id
         """
 
         if self.data_release_date is not None:
@@ -356,7 +359,12 @@ def _get_ophys_experiment_table(self) -> pd.DataFrame:
         query += where_clause
 
         self.logger.debug(f"get_ophys_experiment_table query: \n{query}")
-        return self.lims_engine.select(query)
+        query_df = self.lims_engine.select(query)
+        targeted_imaging_depth = query_df[
+            ["ophys_container_id", "imaging_depth"]
+            ].groupby("ophys_container_id").mean()
+        targeted_imaging_depth.columns = ['targeted_imaging_depth']
+        return query_df.merge(targeted_imaging_depth, on='ophys_container_id')
 
     def _get_ophys_cells_table(self):
         """
@@ -416,7 +424,7 @@ def _get_ophys_session_table(self) -> pd.DataFrame:
         """Helper function for easier testing.
         Return a pd.Dataframe table with all ophys_session_ids and relevant
         metadata.
-        Return columns: ophys_session_id, behavior_session_id,
+        Return columns: ophys_session_id, behavior_session_id, project_code,
                         ophys_experiment_id, project_code, session_name,
                         date_of_acquisition,
                         specimen_id, full_genotype, sex, age_in_days,
@@ -428,10 +436,10 @@ def _get_ophys_session_table(self) -> pd.DataFrame:
             SELECT
                 os.id as ophys_session_id,
                 bs.id as behavior_session_id,
+                pr.code as project_code,
                 imaging_plane_group_count.imaging_plane_group_count,
                 exp_ids.experiment_ids as ophys_experiment_id,
                 cntr_ids.container_ids as ophys_container_id,
-                pr.code as project_code,
                 os.name as session_name,
                 os.date_of_acquisition,
                 os.specimen_id
@@ -528,7 +536,7 @@ def get_behavior_session_table(self, n_workers: int = 1) -> pd.DataFrame:
         else:
             session_metadata = [
                 BehaviorMetadata.from_lims(
-                    behavior_session_id=behavior_session_id,
+                    behavior_session_id=BehaviorSessionId(behavior_session_id),
                     lims_db=db_connection_creator(
                         fallback_credentials=LIMS_DB_CREDENTIAL_MAP
                     )

allensdk/brain_observatory/behavior/behavior_project_cache/tables/ophys_mixin.py

@@ -1,3 +1,6 @@
+import warnings
+
+
 class OphysMixin:
     """A mixin class for ophys project data"""
     def __init__(self):
@@ -11,3 +14,46 @@ def __init__(self):
             self._df = self._df.drop(
                 ['date_of_acquisition_behavior',
                  'date_of_acquisition_ophys'], axis=1)
+        self._clean_up_project_code()
+
+    def _clean_up_project_code(self):
+        """Remove duplicate project_code columns from the data frames. This is
+        as depending on the table we get the project_code either through the
+        behavior_sessions or ophys_sessions tables.
+
+        Additionally test that the values in the columns are identical.
+        """
+
+        if 'project_code_behavior' in self._df and \
+                'project_code_ophys' in self._df:
+
+            if (self._df['project_code_ophys'].isna()).sum() == 0:
+                # If there are no missing ophys_session_ids for the table then
+                # we are loading of the ophys tables and should be able to
+                # compare the ids directly to assure ourselves that the
+                # project ids are the same between ophys and behavior sessions.
+                if not self._df['project_code_ophys'].equals(
+                        self._df['project_code_behavior']):
+                    warnings.warn("BehaviorSession and OphysSession "
+                                  "project_code's do not agree. This is "
+                                  "likely due to issues with the data in "
+                                  "LIMS. Using OphysSession project_code.")
+                self._df['project_code'] = self._df['project_code_ophys']
+            else:
+                # If there are missing ophys_session_ids for the table then
+                # we are loading of the behavior table first will need to mask
+                # to only the sessions that have ophys_session_ids before
+                # comparing project_codes.
+                has_ophys_session_mask = ~self._df['project_code_ophys'].isna()
+                if not self._df['project_code_behavior'][
+                        has_ophys_session_mask].equals(
+                            self._df['project_code_ophys'][
+                                has_ophys_session_mask]):
+                    warnings.warn("BehaviorSession and OphysSession "
+                                  "project_codes do not agree. This is likely "
+                                  "due to issues with the data in LIMS. Using "
+                                  "BehaviorSession project_code.")
+                self._df['project_code'] = self._df['project_code_behavior']
+            self._df = self._df.drop(
+                ['project_code_ophys',  'project_code_behavior'],
+                axis=1)