I'm seeing some weird behaviour with querying a partition table
(partitioned by month) This table has 414 columns, so each partition (if
we select all columns is about 25 GB). This is a special table (not
dealing with this many columns regularly). ra...
From KDB 4.0 release notes: The -w limit (M1/m2) is no longer
thread-local, but domain-local: command-line option -w and system
command \w set the limit for domain 0. Is there a way to circumvent this
so that memory is still thread local in KDB 4.0?
I've noticed some weird behaviour with amend entire using lists of
lists. The behaviour is different depending on whether on I use a file
handle or pass the list directly to the function. Code to recreate is
pasted below: weirdPartition:(1 2 3f;();en...
The .Q namespace – tools | Reference | kdb+ and q documentation - Kdb+
and q documentation (kx.com) based on this, it seems .q.en reads the
sym file from disk every time it is called, which would pick up any
updates that are coming from other proce...
Thank you! Another related question - after doing the initial query on
the partition, when I do 100#rawData That still worked (and was fast,
and did not increase the memory usage). If you're saying we are not
actually loading the partition into memor...
Thank you @megan_mcp ! the lsq won't fit our use case since we want the
coefficients to be strictly greater than zero. We'll have a look at
embedPy - was just hoping somebody had possibly written a native kdb
The weird thing for me is we have a process that writes in this fashion,
running several times a day. In some cases, it fails (which makes sense
given the explanation you have given), but it does work in some cases as
well. might lean towards just re...
What precisely does at a process level mean? Does that mean only if a
single port is multithreading (i.e. peaching and writing to the sym file
the lock will be used)? Or I have two distinct ports both writing to the
sym file, this built in locking wo...