-
Hi Ruoqi, I found the paper to be interested, and thank you for making the codes available. Before I planned on applying the model to our real data problem, I was running your code with the simulation…
-
A colleage wrote this email to me a week ago (I've edited it for clarity):
> Did you ever estimate a synthetic control method (without exogenous X's)? I am currently using the Synth package in R de…
-
-
Hello,
Thank you very much for providing this great method!
I tried to use xfuse on synthetic data, but I ran into some difficulties. How does the synthetic data look like? And how can I modify th…
-
**Is your feature request related to a problem? Please describe.**
Uber/Ludwig wishes to enhance its support for time series modeling and analysis. While there are many existing time series available…
-
Hi @ylabbe ,
I followed your script and tried to generate the synthetic data by running:
> python -m cosypose.scripts.run_dataset_recording --config ycbv --local
However, it showed me an err…
-
https://github.com/lk-geimfari/elizabeth/
-
## Problem Definition
Execution effort weights need recalibration.
Initial calibration was done with [Flip 753](https://github.com/onflow/flips/blob/main/protocol/20220111-execution-effort.md#f…
-
Aren't there explicit tables among the tables?
For example, in the case of hotel-guest data to test the HMA algorithm in SDV, when the hotel data is explicitly used, how can REaLTabFormer synthesiz…
-
For the b-value tests, we are generating random synthetic data. However, this makes the tests slow and prone to error (even if the function is correct, sometimes the tests will fail)
Probably, we s…