Numalogic

* [8dee66b](https://github.com/numaproj/numalogic/commit/8dee66bc29e9799c39da656a3beed26c28242c99) feat!: support static threshold score adjustment (#350)

Contributors

* Avik Basu

* [4e93088](https://github.com/numaproj/numalogic/commit/4e93088c6694457c431fb32e20d5d4477c4579cf) bump dev version
* [81971dc](https://github.com/numaproj/numalogic/commit/81971dcf5045f7e6350067ce73ee84609b7b810a) fix: static threshold
* [db943a6](https://github.com/numaproj/numalogic/commit/db943a6768136bd61bad1cb6614a68680c10f1ec) fix: conditional fwd tags
* [c0225a9](https://github.com/numaproj/numalogic/commit/c0225a967b08e360c426fe35f886c129f2a50363) refactor
* [8f8a532](https://github.com/numaproj/numalogic/commit/8f8a532c3312048489834b570e25c7b8773ebd20) feat: score adjustment with joined trainer vtx
* [ac0eda1](https://github.com/numaproj/numalogic/commit/ac0eda176187c41e4b1aa960b8d0d1cfaeacda38) more tests
* [6623e2b](https://github.com/numaproj/numalogic/commit/6623e2b1d0f85180c472037809b01b4176549eeb) feat: trainer join vertex for preprocess & inference
* [fb97b3a](https://github.com/numaproj/numalogic/commit/fb97b3a5bbe2668bb19a6680c04a6f337f3b05fa) feat: allow chunked druid fetch (#349)
* [d8bda0e](https://github.com/numaproj/numalogic/commit/d8bda0e159cc89af95daa70b2acd6b276a4f49c1) feat: Multivariate Anomaly Detection (#343)
* [d0dcb47](https://github.com/numaproj/numalogic/commit/d0dcb472fe5fcb80df33c51088332f946180e55a) feat: score adjuster using static threshold

Contributors

* Avik Basu

* [b41a7f7](https://github.com/numaproj/numalogic/commit/b41a7f72ca92bd635e20d2cb125d45329038448a) fix: comments
* [780f2e2](https://github.com/numaproj/numalogic/commit/780f2e23cccf3648b385499c589347c3bed0461e) fix merge issues
* [9de1e83](https://github.com/numaproj/numalogic/commit/9de1e83f133b759431ddbf5ab690cd7d0cfca5d3) fix: backtest
* [c418b5a](https://github.com/numaproj/numalogic/commit/c418b5aed2807bc1d8fa3f4a943dd360e998b10e) tmpfix: numalogic opex tags for query filter
* [bc6d08c](https://github.com/numaproj/numalogic/commit/bc6d08c4abf6626a89c2b7555d486c12d24dd301) feat: Score config in postprocess
* [a913545](https://github.com/numaproj/numalogic/commit/a913545c1d1e2197d2ce65714f675a702dc56226) fix: get item for final score
* [cc681a3](https://github.com/numaproj/numalogic/commit/cc681a39fc644ffc5d69e717f291dd1a9e304ce7) tmp: testing per feature score
* [634b2cf](https://github.com/numaproj/numalogic/commit/634b2cf5e3acf1236f30fb4dfdf25578161b5e64) fix: factory class
* [36db5df](https://github.com/numaproj/numalogic/commit/36db5df3bf5c5bb0ebe567445028f9ddd2ab9c1d) tests: median thresh tests
* [777b5e2](https://github.com/numaproj/numalogic/commit/777b5e2415b1e3177e780c1d6806b4d14f77cef1) fix: import
* [a075761](https://github.com/numaproj/numalogic/commit/a075761c4b5cf730ba21f457716a0c5f22cf2d88) feat: add percentile thresholding
* [fbf171f](https://github.com/numaproj/numalogic/commit/fbf171f3b4b39831a0ee7698ed55ca3425b9398f) fix: pipeline id keyerr in tools.py
* [add642c](https://github.com/numaproj/numalogic/commit/add642c3b7ab49f8877fa732c2e6e7362bd25df8) fix: pipeline id key error
* [dc35c80](https://github.com/numaproj/numalogic/commit/dc35c80d26979fd27129405c98557c3aa415106a) try: better docker workflow
* [faf1e32](https://github.com/numaproj/numalogic/commit/faf1e32da4b4ed99a8b1b70acdc905ab97aeccb2) fix: default ml pipe id in preproc
* [03faf8f](https://github.com/numaproj/numalogic/commit/03faf8f723556a074703ca914ceaa28f92289f81) fix: vae nsamples
* [16c02f1](https://github.com/numaproj/numalogic/commit/16c02f158e701a95d1d18e663e83ea19581dd393) feat: add beta parameter for disentanglement
* [5ec5883](https://github.com/numaproj/numalogic/commit/5ec5883e2a3dcfd1df8f477c3cbd8c2be3459e34) feat: difference transform
* [46363d3](https://github.com/numaproj/numalogic/commit/46363d3c0ab01d7a6a6c157ffb9beff294850fa5) feat: output threshold + final scores
* [7f9317c](https://github.com/numaproj/numalogic/commit/7f9317c60209b5f463ecaadb328bb985a7051783) feat: support loading nl conf
* [3035407](https://github.com/numaproj/numalogic/commit/303540770d7452b864df591495d54ae6858c1f16) feat: multivariate backtesting
* [c0596d7](https://github.com/numaproj/numalogic/commit/c0596d7828fcface27fe6ce4bfbc5ece47de1035) add more logs
* [3ca8c15](https://github.com/numaproj/numalogic/commit/3ca8c152a6d35afb69bdbe9dfc872c8de33c3c9a) tmp: try exp mov average
* [43c1ec8](https://github.com/numaproj/numalogic/commit/43c1ec81ed69eeb6852fd56bf51232d00708cbee) feat: add transforms and robust thresholding
* [19ddc8e](https://github.com/numaproj/numalogic/commit/19ddc8ead61e1b720d69b8556385fabfb2746369) Demo branch (#335)

Contributors

* Avik Basu
* Kushal Batra

* [4dbde41](https://github.com/numaproj/numalogic/commit/4dbde415ed214fb984fd3989775ee40c9f7b6bf9) Fix: Get metrics from _conf.metrics on Trainer to avoid issue with Flattening the matrix (#346)

Contributors

* shashank10456

* [824a55d](https://github.com/numaproj/numalogic/commit/824a55dc7c755dedc11ade4b98a0b1c2c1acd6b1) fix: add FlattenVector transformer (#344)
* [146ec00](https://github.com/numaproj/numalogic/commit/146ec0090a87c2ded291a229573e215c556ca0f8) fix: add unified conf (#342)
* [870f263](https://github.com/numaproj/numalogic/commit/870f263166929c904e3c37e7f3ae530a378a963c) fix: take mean before calculating the thresholds (#340)
* [dbb510f](https://github.com/numaproj/numalogic/commit/dbb510fa0ba8e0696200d4aa62b046612db7527b) fix: add max value map for cliping the value (#339)
* [80ef431](https://github.com/numaproj/numalogic/commit/80ef43170bc119369aad6661ffa4a8417ba2b063) "Source" tag for metrics (#338)
* [b292553](https://github.com/numaproj/numalogic/commit/b292553adaa77e6b09731625a9e04849f7651704) Add pl conf (#336)
* [28fa28f](https://github.com/numaproj/numalogic/commit/28fa28fa972e347548f3b0d56e52605abc29f0d1) Metrics (#322)

Contributors

* Kushal Batra

* [146ec00](https://github.com/numaproj/numalogic/commit/146ec0090a87c2ded291a229573e215c556ca0f8) fix: add unified conf (#342)
* [870f263](https://github.com/numaproj/numalogic/commit/870f263166929c904e3c37e7f3ae530a378a963c) fix: take mean before calculating the thresholds (#340)
* [dbb510f](https://github.com/numaproj/numalogic/commit/dbb510fa0ba8e0696200d4aa62b046612db7527b) fix: add max value map for cliping the value (#339)
* [80ef431](https://github.com/numaproj/numalogic/commit/80ef43170bc119369aad6661ffa4a8417ba2b063) "Source" tag for metrics (#338)
* [b292553](https://github.com/numaproj/numalogic/commit/b292553adaa77e6b09731625a9e04849f7651704) Add pl conf (#336)
* [28fa28f](https://github.com/numaproj/numalogic/commit/28fa28fa972e347548f3b0d56e52605abc29f0d1) Metrics (#322)
* [dfc383a](https://github.com/numaproj/numalogic/commit/dfc383a37ee0c950ff01cb9559427f8a13757e60) feat: support both base conf and app conf (#328)
* [8b7f45f](https://github.com/numaproj/numalogic/commit/8b7f45f01427a2614f66d39a643c2c06d423f217) feat!: support full multivariate prometheus fetching (#325)
* [c967f20](https://github.com/numaproj/numalogic/commit/c967f200ddf757d239c65fd97c488023e5daf2ad) feat: support multivar prom (#317)
* [46cdfcd](https://github.com/numaproj/numalogic/commit/46cdfcdd42624263ff796f09ef3b8f20e8bb36a3) add retrain logic for insufficient data (#321)
* [0bbb53d](https://github.com/numaproj/numalogic/commit/0bbb53d63d300c20c545a107257bf78603520013) doc: Update to fix examples ipynb files (#318)
* [61f4575](https://github.com/numaproj/numalogic/commit/61f4575eb941e178d4ad128e870983ff026a78bf) chore!: unify and refactor trainer (#315)

Contributors

* Avik Basu
* Haripriya
* Kushal Batra