Kristin Cowalcijk
Kristin Cowalcijk
Hi, I created a docker images for building manylinux2010 wheels for FMM and integrated it with continuous integration. These wheels should be easily installed on many linux platforms, and it...
This patch allows `Network` to be constructed from scratching by calling `add_edge`, which makes experimenting with various kinds of road network data much easier.
Reverse movement penalties were added to reverse edges. This is for prevending reverse edges from being incorrectly matched. An example for reproducing this issue was added to newly added STMATCH...
Jupyter notebooks were rerun on python3 kernel, using fmm package built by #210 .
Signed-off-by: Kristin Cowalcijk
Vectorized read for parquet was enabled by default since [commit ec2f1ad7](https://github.com/apache/iceberg/commit/ec2f1ad76d2152d6aa08b185be4a2447319db642), this patch updates the doc accordingly.
### Describe the bug I am running a modified version of TPC-H query 10. I've removed the filters to stress comet and see how it behaves when processing large amount...
### What is the problem the feature request solves? The DataFusion Comet documentation has a [memory tuning section](https://datafusion.apache.org/comet/user-guide/tuning.html#memory-tuning) in the tuning guide after addressing https://github.com/apache/datafusion-comet/issues/595, it looks simple at first...
### Describe the bug We've seen this exception when running queries with `spark.comet.exec.shuffle.mode=native`: ``` Py4JJavaError: An error occurred while calling o456.collectToPython. : org.apache.spark.SparkException: Job aborted due to stage failure: Task...
### Describe the bug We easily run into this problem when running queries with `spark.comet.exec.shuffle.mode=jvm`: ``` org.apache.spark.memory.SparkOutOfMemoryError: Unable to acquire 67108848 bytes of memory, got 65700208 bytes. Available: 65700208 org.apache.spark.memory.SparkOutOfMemoryError:...