search

WenjieDu / SAITS

The official PyTorch implementation of the paper "SAITS: Self-Attention-based Imputation for Time Series". A fast and state-of-the-art (SOTA) deep-learning neural network model for efficient time-series imputation (impute multivariate incomplete time series containing NaN missing data/values with machine learning). https://arxiv.org/abs/2202.08516
https://doi.org/10.1016/j.eswa.2023.119619
MIT License
292 stars 48 forks source link

Is this X_tilde_3 or X_c? #26

Closed lzylzylllll closed 1 year ago

lzylzylllll commented 1 year ago 
        imputation_MAE = masked_mae_cal(
            X_tilde_3, inputs["X_holdout"], inputs["indicating_mask"]
        )
WenjieDu commented 1 year ago

Hi there,

Thank you so much for your attention to SAITS! If you find SAITS is helpful to your work, please star⭐️ this repository. Your star is your recognition, which can let others notice SAITS. It matters and is definitely a kind of contribution.

I have received your message and will respond ASAP. Thank you again for your patience! 😃

Best,
Wenjie

WenjieDu commented 1 year ago

Yeah, I know this. No worries. The parts in X_tilde_3 and X_c for this loss calculation are exactly the same.

lzylzylllll commented 1 year ago

Do you mean X_tilde_3 and X_c values are equal?

WenjieDu commented 1 year ago

Of course they are not equal, there is code here to generate X_c from X_tilde_3 https://github.com/WenjieDu/SAITS/blob/main/modeling/saits.py#L193-L196. I mean the exactly what I said above the parts in X_tilde_3 and X_c for this loss calculation are exactly the same. You can replace X_tilde_3 with X_c in that line of code calculating imputation_MAE. The result will be the same.

lzylzylllll commented 1 year ago

I understand, thanks for the answer

WenjieDu commented 1 year ago

Absolutely my pleasure. If it doesn't bother you, please star 🌟 this repo to help more people notice this useful work. Thanks.

WenjieDu commented 1 year ago

I'm closing this issue due to all questions solved. BTW, if you're interested in time series modeling, PyPOTS may be useful to you. Please pay a visit to https://pypots.com/ to know more about it. 😊