Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/3428
Hello @bowenli86, AFAIK nobody's working on the PR's right now. I'm
currently 100% focused on finishing my PhD so I won't have time to review.
---
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/3428
Thank you for the info!
In that case it's a bit ambiguous now which issue this PR is solving. If
you are introducing an interface for GLMs we should be having this discussion
under
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/3428
Hello @mtunique and thank you for you contribution!
Could you briefly describe your changes/additions and how you have tested
the implementation for correctness?
---
If your project is
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/3313
Hello @skonto thanks for your contribution!
I'm currently snowed under paper deadlines, so I can't give you a time for
when I'll be able to go through this, hopefully with
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/1849
Hello @gaborhermann. Personally I prefer to have PRs be as specific as
possible, so I would recommend we try to get this merged before #2838, and then
rebase that on master.
Given the
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1849#discussion_r96406269
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/evaluation/Score.scala
---
@@ -0,0 +1,145 @@
+/*
+ * Licensed to the Apache
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/1849
Hello @skonto this PR will probably be subsumed by #2838, you can check out
the latest development there.
---
If your project is set up for it, you can reply to this email and have your
reply
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/757
Sure @sachingoel0101 feel free to split up the PRs to reduce overhead.
For added initialization schemes let me throw [this recent
NIPS](https://papers.nips.cc/paper/6478-fast-and-provably
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2761
Hello @ch33hau, sorry for the late reply, I've been at a conference the
past week. With the latest changes this LGTM, I've edited the fix version in
JIRA to 1.2.0 to give this more visi
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r91119867
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/SessionWindowing.scala
---
@@ -60,11
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r91119955
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/SessionWindowing.scala
---
@@ -77,9
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r91119626
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/GroupedProcessingTimeWindowExample.scala
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r91119469
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/ml/IncrementalLearningSkeleton.scala
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89645498
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/kafka/ReadFromKafka.scala
---
@@ -0,0 +1,72
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2838
> The problem is not with the evaluate(test: TestType): DataSet[Double] but
rather with evaluate(test: TestType): DataSet[(Prediction,Prediction)].
Completely agree there, I advocated
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89471075
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/GroupedProcessingTimeWindowExample.scala
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89469773
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/ml/IncrementalLearningSkeleton.scala
---
@@ -0,0
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89465547
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/kafka/WriteIntoKafka.scala
---
@@ -0,0 +1,81
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89464880
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/kafka/ReadFromKafka.scala
---
@@ -0,0 +1,72
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89472254
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/WindowWordCount.scala
---
@@ -0,0
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89471865
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/windowing/SessionWindowing.scala
---
@@ -0,0
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2761#discussion_r89469371
--- Diff:
flink-examples/flink-examples-streaming/src/main/scala/org/apache/flink/streaming/scala/examples/ml/IncrementalLearningSkeleton.scala
---
@@ -0,0
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2764
Hello @NicoK , do you think you can include
[FLINK-5087](https://issues.apache.org/jira/browse/FLINK-5087) in this PR, or
should we create a new one?
---
If your project is set up for it, you can
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2819
Hello @gaborhermann,
I really like the idea of introducing a `MatrixFactorization` interface
that we can then use for different specialized optimization algorithms. For the
question I&#
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2838
Hello Gabor,
I like the idea of having a RankingScore, it seems like having that
hierarchy with Score, RankingScore and PairWiseScore gives us the flexibility
we need to include ranking
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2838#discussion_r89292988
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/pipeline/Predictor.scala
---
@@ -72,14 +77,142 @@ trait Predictor[Self] extends
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2740
Hello @tfournier314, I should have clarified for documentation I meant
apart from the docstrings you have added now, we also have to include
documentation in the Flink
[docs](https://github.com
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2838
Hello @gaborhermann, thanks for making the PR!
I'll try to take a look this week, I've been busy with a couple of other
PRs these days.
---
If your project is set up for i
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2740
Hello @tfournier314,
This PR is still missing documentation. After that is done a project
committer will have to review it before it gets merged, which might take a
while.
Regards
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2764
IIRC @vasia is using Eclipse to develop, maybe she can chime in.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2761
Hello Lim, thank you for your contribution!
I've taken a quick look and most of these look fine, plus I see you've
included the required tests.
I'll do a review this wee
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2740
@greghogan Excuse my ignorance, I'm only now learning about Flink internals
:)
It seems like the issue here was that `partitionByRange` partitions keys in
ascending order but we want th
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2740
Hello @tfournier314 I tested your code and it does seem that partitions are
sorted
only internally, which is expected and `zipWithIndex` is AFAIK unaware of
the sorted (as in key range) order
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87422110
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -273,6 +308,14 @@ object ALS {
val
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87421513
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -675,7 +756,69 @@ object ALS
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87202604
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -675,7 +756,69 @@ object ALS
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87199446
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -273,6 +308,14 @@ object ALS {
val
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87195483
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -156,6 +171,26 @@ class ALS extends Predictor[ALS
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r87201508
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -535,8 +581,17 @@ object ALS {
itemOut
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2740#discussion_r87192044
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/preprocessing/StringIndexer.scala
---
@@ -0,0 +1,108 @@
+package
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2542
@gaborhermann Yup the approach taken by the Spark community for testing is
closer to what we would like to have for non-deterministic algorithms, but what
you have implemented now should suffice on
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2542
Hello @gaborhermann,
Yes I think you are right in that respect, just wanted to note that we
should perform some comparative benchmarks in the future.
So the benchmarks look good
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2542
Thank you @jfeher!
Could you clarify what you mean by filtering the data to get unique
item-user pairs? Is this because the iALS algorithm only supports binary
interactions (i.e. number of
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2684
Hello @tillrohrmann I think you make some valid points. My original idea
was to move completely to using `LabeledVector` to the de facto format for
supervised learning problems and do away with
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2735
Thank you for your contribution Kalman!
I just took a brief look, this is a big PR so will probably take some time
to review.
For now a few things that jump to mind
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2682
Thank you @ch33hau I'll also mark the issue as a duplicate!
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project doe
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2684
One last thing, could you change this PR's title to `[FLINK-4865] [ml] Add
EvaluateDataSet operation for LabeledVector`
This is the standard naming for PRs, issue number, follow
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2684
> We don't need to care about Jenkins/Travis fails, do we ?
Normally we should, but there many unstable tests and issues with Travis
that it's hard to follow now.
You sho
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2684
Thank you for your contribution Thomas!
The changes look mostly fine I think, my only concern is the need to cast
the types in lines
[215-216](https://github.com/apache/flink/pull/2684
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2668
Hello Thomas, thank you for your contribution!
I took a brief look so some initial comments:
This seems to be making changes to `MLUtils` which AFAIK is outside the
scope of this
GitHub user thvasilo opened a pull request:
https://github.com/apache/flink/pull/2658
[FLINK-4850] [ml] FlinkML - SVM predict Operation for Vector and not
LaveledVector
The current version of the quickstart guide includes erroneous code, this
changes the function calls to have
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/1849
@gaborhermann In terms of missing features, documentation is definitely
missing, as @rawkintrevo mentioned.
For the issues mentioned in the JIRA issue you linked I've replied on the
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2542
@gaborhermann I think have a larger scale test would boost our confidence
in the implementation, and maybe point out some problems that do not manifest
with small data, which is very common
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r81159482
--- Diff:
flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/recommendation/ImplicitALSTest.scala
---
@@ -0,0 +1,171 @@
+/*
+ * Licensed
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r81159171
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -581,6 +637,16 @@ object ALS {
val
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80860723
--- Diff: docs/dev/libs/ml/als.md ---
@@ -99,6 +114,26 @@ The alternating least squares implementation can be
controlled by the following
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80862241
--- Diff:
flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/recommendation/ImplicitALSTest.scala
---
@@ -0,0 +1,171 @@
+/*
+ * Licensed
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80861067
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/recommendation/ALS.scala
---
@@ -581,6 +637,16 @@ object ALS {
val
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80862018
--- Diff:
flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/recommendation/ImplicitALSTest.scala
---
@@ -0,0 +1,171 @@
+/*
+ * Licensed
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80860057
--- Diff: docs/dev/libs/ml/als.md ---
@@ -49,6 +49,21 @@ By applying this step alternately to the matrices $U$
and $V$, we can iterativel
The
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80860320
--- Diff: docs/dev/libs/ml/als.md ---
@@ -99,6 +114,26 @@ The alternating least squares implementation can be
controlled by the following
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/2542
Hello @gaborhermann thank you for your contribution!
Are the numbers here non-zero entries in a matrix?
If that is the case do you think it would be possible to test this on some
larger
GitHub user thvasilo opened a pull request:
https://github.com/apache/flink/pull/2393
[trivial] Fix typo in dosctring
There is a small typo in the getBufferTimeout docstring, this fixes it.
Talked with @tillrohrmann about whether it's worth it to open a PR for
some
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/1220
Oh, sorry should have checked the master docs.
On Jul 26, 2016 5:29 AM, "Chiwan Park" wrote:
> Hi @thvasilo <https://github.com/thvasilo>, I've checked the
Github user thvasilo commented on the issue:
https://github.com/apache/flink/pull/1220
Hello Daniel, sorry to bring this up months later,but I see that while the
documentation exists, there is nothing linking to it from the FlinkML index
page. Would you care to create a new PR
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1985#discussion_r63519957
--- Diff:
flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/optimization/RegularizationPenaltyITSuite.scala
---
@@ -0,0 +1,65
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1985#discussion_r63519390
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/optimization/RegularizationPenalty.scala
---
@@ -0,0 +1,215
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1985#discussion_r63517610
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/optimization/GradientDescent.scala
---
@@ -272,7 +272,7 @@ abstract class
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1985#discussion_r63517587
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/optimization/GradientDescent.scala
---
@@ -272,7 +272,7 @@ abstract class
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/656#issuecomment-218399619
Hello @skavulya thank you for your contribution, I'm looking forward to the
PR!
Re. the regularization penalty, we it changed to make user choice easier,
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1915#issuecomment-215201773
Any idea what might be causing the error in FlinkML? I can't pinpoint it to
any specific source file.
---
If your project is set up for it, you can reply to
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1849#issuecomment-212977748
I did some testing and I think the problem has to do with the types that
each scaler expects.
`StandardScaler` has fit and transform operations for `DataSets
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1849#issuecomment-212942014
Well breeze was recently bumped to 0.12 #1876, maybe that has something to
do with it, but let's see.
Any chance you can try with the prev. Breeze ve
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1849#issuecomment-212793430
Hello Trevor,
Thanks for taking the time to look at this, I'll investigate these issues
today hopefully.
--
Sent from a mobile device
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1849#issuecomment-205214498
@mbalassi @tillrohrmann
Closed the previous PR and opened this one for the evaluation framework, as
I had some issues with rebasing.
---
If your project is
Github user thvasilo closed the pull request at:
https://github.com/apache/flink/pull/871
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/871#issuecomment-205214045
Moving this to #1849 due to git eating my changes when rebasing under the
new dir structure.
---
If your project is set up for it, you can reply to this email and have
GitHub user thvasilo opened a pull request:
https://github.com/apache/flink/pull/1849
[FLINK-2157] [ml] Create evaluation framework for ML library
Using this PR instead of #871 due to rebase issues.
You can merge this pull request into a Git repository by running:
$ git pull
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/871#issuecomment-203075630
@rawkintrevo AFAIK it's lack of time from a commiter to review it. If
@tillrohrmann can find some time to review this I'll refactor it to get rid of
the con
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1688#issuecomment-188278485
In terms of usability I think it is.The whole point of the quickstart guide
is to get users up and running as quickly as possible. For their first toy
project people
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/1688#issuecomment-188258580
Thanks for this and fixing the SBT build, this should greatly help setting
up new users in demonstrations etc.
@tillrohrmann Any reason giter8 was selected as
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1587#discussion_r51910120
--- Diff:
flink-libraries/flink-ml/src/main/scala/org/apache/flink/ml/optimization/GradientDescent.scala
---
@@ -192,10 +190,18 @@ abstract class
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/1587#discussion_r51909968
--- Diff:
flink-libraries/flink-ml/src/test/scala/org/apache/flink/ml/regression/RegressionData.scala
---
@@ -27,6 +27,21 @@ object RegressionData
GitHub user thvasilo opened a pull request:
https://github.com/apache/flink/pull/1100
[FLINK-2003] [docs] Building on some encrypted filesystems leads to "File
name too long" error
Replaces #690, adding docs instead of changing the pom files.
You can merge this pull requ
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137936053
I agree, let's not break the API in this PR. We can create an issue and
have a small discussion on the list about the change, and if the community
agrees we can
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-13090
If it is actually considered to be API breaking then we should open a new
PR and JIRA to make sure the change is well documented. Let's see what Till
thinks and
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137775231
Also @tillrohrmann : Is changing the DataSetUtils structure considered an
API breaking change and should we handle it differently?
Any code using `import
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137772917
I think it would be hard to do otherwise. `java.lang.Double`'s are objects
but `scala.Double`'s are not from what I understand. In this case duplicating
the
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137768373
Re 1: Good change, implicit classes is the way to go here, no need for
implicit conversion in the object.
---
If your project is set up for it, you can reply to this
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137764885
```scala
def createDiscreteHistogram: DataSet[DiscreteHistogram] = {
wrap(jutils.DataSetUtils.createDiscreteHistogram(
self.map(x =>
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137759354
@sachingoel0101 That sounds like a good place for the docs, that
`zip_elements_guide.mb` is linked from the DataSet Transformations doc, so
replacing that link to a
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-137714631
Hello, my 2c: This PR should include docs and a wrapper for the Scala API.
We can also do this with a separate issue but it would be best if we merge as a
more complete
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/949#issuecomment-133675023
It's great to have this in, I'll try to update the cross-validation and SGD
to use this.
---
If your project is set up for it, you can reply to this emai
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-132552269
Sounds good, just make sure to update the title and description of both PRs
to reflect the current state
---
If your project is set up for it, you can reply to this
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-132261492
Sorry to be a nitpick, but if we are going to split the PR then the
documentation should be split accordingly, we can merge the column-wise
statistics once this one is
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-132192126
Hello, I was wondering when the column-wise statistics were added to this
PR and whether it makes sense to add them here, or create a new issue and PR
for them
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-132146208
Sounds good.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/861#issuecomment-132131366
I would vote for returning a continuous histogram as the default and
perhaps provide a discrete one as private to the ml package.
---
If your project is set up for it
Github user thvasilo commented on the pull request:
https://github.com/apache/flink/pull/696#issuecomment-131107449
+1 for closing this and focusing on approximate kNN instead.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as
Github user thvasilo commented on a diff in the pull request:
https://github.com/apache/flink/pull/949#discussion_r37054584
--- Diff:
flink-tests/src/test/scala/org/apache/flink/api/scala/operators/SampleITCase.scala
---
@@ -0,0 +1,166 @@
+/*
+ * Licensed to the Apache
1 - 100 of 395 matches
Mail list logo
