Hi, @Alexey Zinoviev .
Currently, Ignite supports only these distances
(org.apache.ignite.ml.math.distances) :
- ChebyshevDistance
- CosineSimilarity
- EuclideanDistance
- HammingDistance
- JaccardIndex
- ManhattanDistance
- MinkowskiDistance
But in scipy (
https://docs.scipy.org/doc/scipy/refere
Hi guys,
It would be really cool to have async operations.
My point is that the worst thing that we have at the moment is that we need
to write data to the socket under the sndLock.
It makes it nearly impossible for users to implement async operations on
their side.
Of course they can scale up by
Great, I'll create a ticket for that case and assign it for you, I think it
could be very useful contribution.
вт, 25 авг. 2020 г., 10:07 Mark Andreev :
> Hi, @Alexey Zinoviev .
>
> Currently, Ignite supports only these distances
> (org.apache.ignite.ml.math.distances) :
> - ChebyshevDistance
>
Stanilovsky Evgeny created IGNITE-13383:
---
Summary: Java thin client improvements: channel reconnect and
redundant concurrency structures replacement.
Key: IGNITE-13383
URL: https://issues.apache.org/jira/bro
Zhenya,
Stacktraces are considered to be able to expose sensitive information about
code, see [1]. So as previously, I agree with Pavel that we should have an
option to disable this behavior.
[1]
https://wiki.sei.cmu.edu/confluence/display/java/ERR01-J.+Do+not+allow+exceptions+to+expose+sensitive
Alexey.
This option should be on the server side, so server administrator can disable
stack trace for all clients.
Correct?
> 25 авг. 2020 г., в 12:47, Alexey Goncharuk
> написал(а):
>
> Zhenya,
>
> Stacktraces are considered to be able to expose sensitive information about
> code, see [1].
Hello folks,
I agree with Val: IgniteFuture was created long before we could use the
Java futures and is being kept for compatibility reasons. While keeping API
consistent between thin/thick clients is a good reason, I think moving to
Java futures benefits more to the project and end-users.
Agree
>
> Alexey.
>
> This option should be on the server side, so server administrator can
> disable stack trace for all clients.
> Correct?
>
Yes, correct.
Thanks ! now its clear for me, ones more want to repeat my position — sending
exceptions to the client side looks like bad design, because exceptions its
about unhealthy system recognition from administrator side, not from client of
course, but looks like for now it`s more informative than exi
Alexey Kukushkin created IGNITE-13384:
-
Summary: Auto-generated README.txt contains invalid
workingDirectory property name
Key: IGNITE-13384
URL: https://issues.apache.org/jira/browse/IGNITE-13384
I am wrong Alex but let me provide some history: when I originally
developed the Java thin client 3 years ago I suggested to take all public
API out of ignite-core.jar into a separate ignite-api.jar and develop Java
thin client implementing the common API from the ignite-api.jar. So the
ignite-core
Hi Igniters,
Separate API module sounds nice to me as well. And I also think it is
fine that thin client does not support as much operations as client
node, because it is appealing to evolve API not strictly
synchronously. And I can imagine several options:
1. Have rich API with possibility to hav
Thanks! But now, I have another problem, which is I can't log in, because some
time ago I forgot my password and needed to get a new one by e-mail, but it
didn't come. COuld we find any solution to this problem? Maybe you could send a
new password to my account e-mail, or delete my account? In t
Kirill Tkalenko created IGNITE-13385:
Summary: Documentation of cache warm-up strategy
Key: IGNITE-13385
URL: https://issues.apache.org/jira/browse/IGNITE-13385
Project: Ignite
Issue Type
Hi,
You may create an INFRA ticket in JIRA
https://issues.apache.org/jira/projects/INFRA/issues describing the
problem.
As a workaround you can try your gmail address in cwiki. I expect it
might be simpler than using a corporate one.
2020-08-25 10:49 GMT+03:00, lonsdalel lonsdalel :
> Thanks! Bu
Hi, @Alexey Zinoviev .
I suggest to add a final class modifier and to add a private constructor
to Util classes in ignite ml. This is Sonar rule RSPEC-1118 (
https://rules.sonarsource.com/java/tag/design/RSPEC-1118).
Motivation:
Utility classes, which are collections of static members, are not me
Hi, @Alexey Zinoviev .
I found that class Blas (org.apache.ignite.ml.math) is not Serializable but
fields f2jBlas and nativeBlas are transient. So I suggest adding
a Serializable to Blas class.
--
Best regards,
Mark Andreev
Igniters,
As was discussed before, we've scheduled two meetups devoted to Ignite 3.0:
- Sep 15 (in English):
https://www.meetup.com/Apache-Ignite-Virtual-Meetup/events/272675408/
- Sep 17 (in Russian):
https://www.meetup.com/Moscow-Apache-Ignite-Meetup/events/272675398/
I will presen
Hi Val,
Thank you for scheduling the meetups. Will join and connect on Ignite 3.0
release planning.
Regards,
Saikat
On Tue, Aug 25, 2020 at 4:49 PM Valentin Kulichenko <
valentin.kuliche...@gmail.com> wrote:
> Igniters,
>
> As was discussed before, we've scheduled two meetups devoted to Ignite
Here is a sample project I threw together and I've also updated the wiki
just a bit. working on updating the documentation on the micronaut end.
https://cwiki.apache.org/confluence/display/IGNITE/Micronaut+Integration
https://github.com/pollend/micronaut-ignite-sample
On Mon, Aug 24, 2020 at 6:57
20 matches
Mail list logo