Inconsistent epoch naming between TS (E0-E3) and TLS (epoch_1-epoch_4)

#3
by dacom - opened

Hello,

Thank you for sharing this dataset.

While using the dataset, I noticed a potential ambiguity in the epoch naming between the TS and TLS data:

TS data uses E0, E1, E2, E3
TLS data uses epoch_1, epoch_2, epoch_3, epoch_4
From the current documentation, it is not clear whether the correspondence is:

E0 -> epoch_1, E1 -> epoch_2, E2 -> epoch_3, E3 -> epoch_4
or something else.

I checked the dataset documentation files, but I could not find an explicit statement confirming this mapping.

This is important because the TLS documentation describes the deformation sequence as:

epoch_1 -> epoch_2: vertical translation only
epoch_2 -> epoch_3: translation + rotation
epoch_3 -> epoch_4: rotation only
For users who want to compare TS and TLS results across epochs, an explicit mapping between TS epochs and TLS epochs would be very helpful.

Could you please confirm whether:

E0 = epoch_1
E1 = epoch_2
E2 = epoch_3
E3 = epoch_4
If this is correct, it would be great to add it to the dataset documentation to avoid confusion.

Thank you!

Hi,

Thank you for your interest in the dataset and for pointing this out.

You are correct. The mapping (TS β†’ TLS) is:

E0 = epoch_1
E1 = epoch_2
E2 = epoch_3
E3 = epoch_4

We have updated the documentation to make this explicit.

Kind regards,
Zhaoyi

Sign up or log in to comment