Data shape mismatch in ROI segmentation data

[CodyCBakerPhD]

Two PumpProbe sessions are affected by this issue:

• pumpprobe_20210917_132736
• pumpprobe_20211008_115929

Demo script:

import json
import pathlib
import pickle

base_path = pathlib.Path("D:/Leifer")
pump_probe_folder_path = base_path / "20210917" / "pumpprobe_20210917_132736"

signal_file_path = pump_probe_folder_path / "green.pickle"
with open(file=signal_file_path, mode="rb") as io:
    signal_info = pickle.load(file=io)

brains_file_path = pump_probe_folder_path / "brains.json"
with open(brains_file_path, "r") as io:
    brains_info = json.load(fp=io)

assert signal_info.data.shape[1] == len(brains_info["labels"][30]), "Shape mismatch between signal data and ROI labels."

Basically, there are more labeled ROIs (from the NeuroPAL coregistration) than there are tracked ROI fluorescence traces

@emosb Any ideas what might be going on here, or if it is fixable? Otherwise I will skip the processed data for these two sessions