Worrying result after accessing a protocol after some time

Bugs
1

Dear Brainstorm I wish to share something that I've observed which is quite troubling.

I have a protocol investigating the ABRs in a group of people.
I've made custom processes to compute weighted averaging, noise per epoch and residual noise from X number of trials in the ABR acquisition (shared here below).
I haven't accessed this protocol for some time because I had to physically move to a different location with a different computer. Therefore I had to detach the protocol (within an external hard drive) from the previous computer and attach it to a new computer.

Today I ran the same three custom processes again in the same protocol and completely serendipitously found a couple of troubling observations (see data tree below):

  1. The output from the residual noise (RN) and noise per epoch (NEpoch) processes (which are residual noise and noise per epoch files respectively) are different from the existing file.
  2. The output from the weighted averaging process (weighted average file) is the same as the existing file.

This is extremely strange because:

  1. All three custom processes use the same X number of epochs therefore the repeated outputs from these processes should be the same as the already existing files generated from the same processes.
  2. The outputs from the RN and NEpoch processes which are different to the existing files possibly indicate something wrong with the 6000 trials but the weighted average file was the same as the original therefore it isn't a database issue.
  3. If it is not a database issue, then it could be the coding issue, but the custom processes' code did not change at all.
  4. Reloading the database did not solve the issue.

I am confident that my original files generated on 22nd June 2023 are the true results but I am completely baffled as to what is happening here :frowning: . I have attached the link to the bdf file responsible for this data tree and the ABR pipeline I used to generate the 6000 trials.

Many thanks for your time.

CTRL9_L_C60_0_Q.bdf

1691×844 191 KB

ABR_pipeline.mat (6.4 KB)
process_residual_noise.m (8.4 KB)
process_noise_per_epoch.m (4.5 KB)
process_weighted_averaging.m (4.2 KB)

2

Hi @mcp0228,

The shared pipeline file (ABR_pipeline.mat) shows that the process process_wang_rnoise_nepoch.m is used for the computations rather than the individual processes also shared.

Could it be this the origin of the difference?

image

3

Thank you Raymundo, the shared pipeline containing process_wang_rnoise_nepoch.m is the combined process of the individual processes that I've shared.
But I've since abandoned the approach to delve into this further as I don't have time.