Closed Mathux closed 1 year ago
Hi Sebastian,
Thanks for releasing your great work :)
To compute the amplitude, it seems to be a mismatch between the formula in the paper and in the code:
amp = 2 * torch.sqrt(torch.sum(power, dim=dim)) / self.time_range
A = sqrt(2/N * sum(p_ij))
I think either the code should: amp = torch.sqrt(2 / self.time_range * torch.sum(power, dim=dim)) or the paper should be: A = 2 / N * sqrt(sum(p_ij))
amp = torch.sqrt(2 / self.time_range * torch.sum(power, dim=dim))
A = 2 / N * sqrt(sum(p_ij))
Best, Mathis
I just found that the 2/N factor is already defined in pi_j, so the formula is correct!
Hi Sebastian,
Thanks for releasing your great work :)
To compute the amplitude, it seems to be a mismatch between the formula in the paper and in the code:
amp = 2 * torch.sqrt(torch.sum(power, dim=dim)) / self.time_range
A = sqrt(2/N * sum(p_ij))
I think either the code should:
amp = torch.sqrt(2 / self.time_range * torch.sum(power, dim=dim))
or the paper should be:A = 2 / N * sqrt(sum(p_ij))
Best, Mathis