Update stage 5 passed column

src/npc_sessions/sessions.py

@@ -41,6 +41,7 @@
 import pynwb
 import upath
 import zarr
+from DynamicRoutingTask.Analysis import DynamicRoutingAnalysisUtils
 from DynamicRoutingTask.Analysis.DynamicRoutingAnalysisUtils import DynRoutData
 
 import npc_sessions.trials as TaskControl
@@ -1981,6 +1982,7 @@ def is_naive(self) -> bool:
 
     @property
     def is_stage_5_passed(self) -> bool:
+        """Before this session, had the subject passed stage 5 (two sessions meeting crossmodal dprime threshold)?"""
         if self.is_templeton:
             logger.warning(
                 f"{self.id} is a Templeton session: returning is_stage_5_passed = False, but we don't currently track this"
@@ -2002,79 +2004,14 @@ def is_stage_5_passed(self) -> bool:
         df = df[df["task version"].str.startswith("stage 5")]
         if df.empty:
             return False
-
-        # from https://github.com/samgale/DynamicRoutingTask/blob/aa781757b3e14895b50851d71583257e8fc39fb9/Analysis/DynamicRoutingAnalysisUtils.py#L282
-        def getPerformanceStats(df, sessions):
-            hits = []
-            dprimeSame = []
-            dprimeOther = []
-            for i in sessions:
-                if isinstance(df.loc[i, "hits"], str):
-                    hits.append(
-                        [int(s) for s in re.findall("[0-9]+", df.loc[i, "hits"])]
-                    )
-                    dprimeSame.append(
-                        [
-                            float(s)
-                            for s in re.findall(
-                                "-*[0-9].[0-9]*|nan", df.loc[i, "d' same modality"]
-                            )
-                        ]
-                    )
-                    dprimeOther.append(
-                        [
-                            float(s)
-                            for s in re.findall(
-                                "-*[0-9].[0-9]*|nan",
-                                df.loc[i, "d' other modality go stim"],
-                            )
-                        ]
-                    )
-                else:
-                    hits.append(df.loc[i, "hits"])
-                    dprimeSame.append(df.loc[i, "d' same modality"])
-                    dprimeOther.append(df.loc[i, "d' other modality go stim"])
-            return hits, dprimeSame, dprimeOther
-
-        def getSessionsToPass(
-            mouseId, df, sessions, stage, hitThresh=100, dprimeThresh=1.5
-        ):
-            sessionsToPass = np.nan
-            for sessionInd in sessions:
-                if sessionInd > sessions[0]:
-                    hits, dprimeSame, dprimeOther = getPerformanceStats(
-                        df, (sessionInd - 1, sessionInd)
-                    )
-                    if (
-                        stage in (1, 2)
-                        and all(h[0] >= hitThresh for h in hits)
-                        and all(d[0] >= dprimeThresh for d in dprimeSame)
-                    ) or (
-                        stage == 5
-                        and np.all(
-                            np.sum(
-                                (np.array(dprimeSame) >= dprimeThresh)
-                                & (np.array(dprimeOther) >= dprimeThresh),
-                                axis=1,
-                            )
-                            > 3
-                        )
-                    ):
-                        sessionsToPass = np.where(sessions == sessionInd)[0][0] + 1
-                        break
-            if np.isnan(sessionsToPass):
-                if stage in (1, 2) and mouseId in (614910, 684071, 682893):
-                    sessionsToPass = len(sessions)
-            return sessionsToPass
-
         return np.isnan(
-            getSessionsToPass(
+            DynamicRoutingAnalysisUtils.getSessionsToPass(
                 mouseId=int(self.subject.id),
                 df=df,
                 sessions=np.where(
                     [
                         str(d).split(" ")[0]
-                        <= self.session_start_time.strftime("%Y-%m-%d")
+                        < self.session_start_time.strftime("%Y-%m-%d")
                         for d in df["start time"].values
                     ]
                 )[0],