Fit V1 cell

Learning objectives

Learn how to combine GLM with other modeling approach.
Review previous tutorials.

import jax
import math
import os
import matplotlib.pyplot as plt
import nemos as nmo
import numpy as np
import pynapple as nap
import requests
import tqdm
import workshop_utils

# required for second order methods (BFGS, Newton-CG)
jax.config.update("jax_enable_x64", True)
# configure plots some
plt.style.use(workshop_utils.STYLE_FILE)

Data Streaming

path = workshop_utils.data.download_data("m691l1.nwb", "https://osf.io/xesdm/download",
                                         '../data')

Pynapple

data = nap.load_file(path)

print(data)

epochs = data["epochs"]
spikes = data["units"]
stimulus = data["whitenoise"]

stimulus is white noise shown at 40 Hz
white noise is a good stimulus for mapping basic stimulus properties of V1 simple cells

fig, ax = plt.subplots(1, 1, figsize=(12,4))
ax.imshow(stimulus[0], cmap='Greys_r')
stimulus.shape

print(spikes)
spikes = spikes[[34]]

goal is to predict the neuron's response to this white noise stimuli
several ways we could do this, what do you think?

Spike-triggered average

compute spike-triggered average to visualize receptive field.

visualize spike-triggered average and decide on our spatial filter.

use the spike-triggered average to preprocess our visual input.

Preparing data for nemos

get counts and filtered_stimulus into proper shape for nemos

Set up the basis and prepare the temporal predictor for the GLM.

Fitting the GLM

Fit the GLM

Examine the resulting temporal filter

Further exercises

Total running time of the script: ( 0 minutes 0.000 seconds)

Download Python source code: 04_v1_cells_users.py

Download Jupyter notebook: 04_v1_cells_users.ipynb

Gallery generated by mkdocs-gallery