Aadyot Bhatnagar

Hi, thanks for your question @tszumowski. Merlion is already capable of supporting multivariate time series datasets for forecasting and anomaly detection. For forecasting, I suggest you check out `ts_datasets.forecast.SeattleTrail`. For...

Ah, thanks for the clarification. In the paper, we actually train a separate model for each time series. In this case, the data loader is iterable as ``for time_series, metadata...

@tszumowski, @isenilov has added an initial version of multi-series training for the DAGMM model in #65. Does this roughly match your expectation? As we consider adding a more general version...

Please provide a full code example of what you tried, so I can help you debug.

I suspect you'll need to set the environment variable `$JAVA_HOME` appropriately, and ensure that `java`'s bin directory is on your `$PATH`.

Thanks for the comment. For time series cross validation and backtesting, I suggest you look into Merlion's [evaluation framework](https://opensource.salesforce.com/Merlion/merlion.evaluate.html). I believe this should address a lot of your concerns. If...

@anton164 Thanks for the comment. You highlight a fundamental challenge with anomaly detection -- often, ground truth labels are unavailable. But in my experience, the most common metrics people use...

Thanks for clarifying. From an implementation perspective, I'd suggest leaving `TSADEvaluator` unchanged. It would probably be much simpler to just extend `TSADScoreAccumulator`. For the specific metrics you mention, you might...

Thanks for the comment. As far as multi-CPU support goes, the underlying API supports it for some models (Prophet, tree models, random cut forest, isolation forest; refer to API documentation...

> Can we add a unit test please? Is there a good template I can base this test on? I'm not sure how to isolate the impact of this change...