Processed per-frame and swim bout data as well as predicted activity for Gradient Navigation dataset

Dataset

The dataset contains two zip-archives. *_NPY.zip contains numpy array files of per-camera-frame predicted activity and per-bout predicted activity across medullar response types. *_Pickle.zip contains pickled dataframes of per-camera-frame fish position and heading data as well as pickled dataframes of per-swimbout kinematic data. For each fish there is one corresponding _fish and one corresponding _bout dataframe pickle.

Gradient_PredictedNeuralActivity_NPY.zip

For each fish in the gradient experiments there is one file named according to .nwb&&_fish_activity.npy. This file contains an n_camera_frames by 7 matrix with activity across all seven medullar response types at each experimental frame (acquired at 100 Hz). For missing frames (including for the first three minutes of the data) NaN values will be present in the data.

For each fish in the gradient experiments there is one file named according to .nwb&&_bout_activity.npy. This file contains an n_swim_bouts by 21 matrix with activity across all seven medullar response types. Each row contains the activity of each of the 7 response types at the time of the bout, 500 ms into the past and 1000 ms into the past (in this order).

Gradient_ProcessedDataframes_Pickle.zip

For each fish in the gradient experiments there is one file named according to .nwb&&_fish.pkl and one file named according to .nwb&&_bout.pkl. Each of these files is a pickled pandas dataframes (to be loaded via pandas.read_pickle ). The *_fish files contain frame-by-frame information about position, speed and heading while the *_bout files contain per-swim-bout start/end information as well as kinematics and states.

Fish dataframe structure

Temperature
X Position
Y Position
Heading
Raw X
Raw Y
Instant speed

The temperature at the fish position at the given frame
The filtered X position of the fish at the given frame in mm
The filtered Y position of the fish at the given frame in mm
The heading of the fish at the given frame in radians
The raw tracked X Position
The raw tracked Y Position
The instantaneous speed of the fish at the given frame in mm/s

Bout dataframe structure

State
Gradient direction
Original index
Start
Stop
Peak speed
Displacement
Angle change
IBI

The annotated swim mode
The cosine of the angle of the bout vector with respect to the gradient direction
The original index of the bout (since bouts close to the arena edge were filtered out)
The start camera frame of the bout
The end camera frame of the bout
The maximal speed in mm/s reached during the bout
The displacement of the bout in mm
The angle change of the bout in radians
The waiting time since the previous bout (interbout interval) in ms

Temperature
Prev Delta T
1s Delta T
Delta X
Delta Y
Prev angle change
X Position
Y Position
Heading

The temperature at the start of the bout
The temperature change across the previous bout
The temperature change across the preceding second
The movement in the X direction in mm
The movement in the Y direction in mm
The angle change of the previous bout
The x position at the start of the bout
The y position at the start of the bout
The fish heading at the start of the bout