

In [2]:

    
import os
import numpy as np
import scipy.io

from yass.evaluate.visualization import ChristmasPlot
from yass.evaluate.util import main_channels

Create Some Fake Entires That demonstrates Plotting

In the constructor, give a title, number of total datasets that you want to plot side by side, a list of methods for which you are plotting results. logit_y will logit transforms the y axis for emphasis on low and high end part of the metric. eval_type simply is for naming purposes, and will appear in y-axis titles.

In the following block we just create fake SNR and metrics just fo demonstration purposes.

If you want to compute SNR of a templates (np.ndarray of shape (# time samples, # channels, # units)) just call main_channels(templates).



In [11]:

    
plot = ChristmasPlot('Fake', n_dataset=3, methods=['yass', 'kilosort', 'spyking circus'], logit_y=True, eval_type="Accuracy")

for method in plot.methods:
    for i in range(plot.n_dataset):
        x = (np.random.rand(30) - 0.5) * 10
        y = 1 / (1 + np.exp(-x + np.random.rand()))
        plot.add_metric(x, y, dataset_number=i, method_name=method)

Generate SNR vs Metric

Change save_to to a file path to save it to file.



In [12]:

    
plot.generate_snr_metric_plot(save_to=None)

Generate the curve plots

Similar to the other plot, give path to file in save_to for saving to file.



In [14]:

    
plot.generate_curve_plots(save_to=None)



In [ ]: