tygert commented on issue #16556: [SPARK-19184][MLlib] Improve numerical stability for method tallSkinnyQR. URL: https://github.com/apache/spark/pull/16556#issuecomment-531962310 Thanks! I haven't been using Spark for a while now, but the bad rap it has for outputting garbage even when the input is fine might be worth fixing. Bear in mind that the kind of example @hl475 produced is the rule, not the exception: if you perform dimensionality reduction or principal component analysis, then some matrix involved in the analysis is very likely to be numerically ill-conditioned. MLlib currently cannot handle ill-conditioned matrices, which makes its use for dimensionality reduction and principal component analysis highly problematic. We have other tools where I work, but not all people have such luxuries.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
