Mailing-List: contact user-help@flink.apache.org; run by ezmlm
Precedence: bulk
MIME-Version: 1.0
In-Reply-To: <ac720b6815784e6097c742f6b6c2bf35@PLSWE13M05.ad.sprint.com>
References: <ac720b6815784e6097c742f6b6c2bf35@PLSWE13M05.ad.sprint.com>
From: Fabian Hueske <fhueske@gmail.com>
Date: Wed, 19 Jul 2017 21:11:30 +0200
Message-ID: <CAAdrtT2umbUatZBqwbkGhPAiRB_Tmjm+c01vODeOjrE9Xrn_mg@mail.gmail.com>
Subject: Re: Flink ML with DataStream
To: "Branham, Jeremy [IT]" <Jeremy.D.Branham@sprint.com>
Cc: "user@flink.apache.org" <user@flink.apache.org>
Content-Type: multipart/related; boundary="f403045e1d90211ef70554b068a4"
archived-at: Wed, 19 Jul 2017 19:12:06 -0000

--f403045e1d90211ef70554b068a4
Content-Type: multipart/alternative; boundary="f403045e1d90211ef60554b068a3"

--f403045e1d90211ef60554b068a3
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi,

unfortunately, it is not possible to convert a DataStream into a DataSet.
Flink's DataSet and DataStream APIs are distinct APIs that cannot be used
together.

The FlinkML library is only available for the DataSet API.
There is some ongoing work to add a machine learning library for streaming
use cases as well, but this is still in an early stage and mostly focusing
on model serving on streams, i.e, applying an externally trained model on
streaming data.

Best, Fabian


2017-07-19 19:07 GMT+02:00 Branham, Jeremy [IT] <Jeremy.D.Branham@sprint.co=
m
>:

> Hello =E2=80=93
>
> I=E2=80=99ve been successful working with Flink in Java, but have some tr=
ouble
> trying to leverage the ML library, specifically with KNN.
>
> From my understanding, this is easier in Scala [1] so I=E2=80=99ve been c=
onverting
> my code.
>
>
>
> One issue I=E2=80=99ve encountered is =E2=80=93 How do I get a DataSet[Ve=
ctor] from a
> DataStream[MyClass]?
>
> I=E2=80=99ve attempted to use windowing, but scala is completely new to m=
e and I
> may need a push in the right direction.
>
>
>
> The below code executes properly, I=E2=80=99m just unsure of the next ste=
p.
>
>
>
>
>
> I=E2=80=99ve also seen an example [2] that looks like something I need to
> implement =E2=80=93 especially the PartialModelBuilder.
>
> Am I on the right track?
>
> Thoughts?
>
>
>
> Thanks!
>
>
>
>
>
> [1] - https://stackoverflow.com/questions/44039857/is-there-a-
> apache-flink-machine-learning-tutorial-in-java-language/44040819#44040819
>
> [2] - https://github.com/apache/flink/blob/master/flink-
> examples/flink-examples-streaming/src/main/scala/org/
> apache/flink/streaming/scala/examples/ml/IncrementalLearningSkeleton.scal=
a
>
>
>
>
>
>
>
> Jeremy D. Branham
>
> Technology Architect - Sprint
> O: +1 (972) 405-2970 <(972)%20405-2970> | M: +1 (817) 791-1627
> <(817)%20791-1627>
>
> Jeremy.D.Branham@Sprint.com
>
> #gettingbettereveryday
>
>
>
> ------------------------------
>
> This e-mail may contain Sprint proprietary information intended for the
> sole use of the recipient(s). Any use by others is prohibited. If you are
> not the intended recipient, please contact the sender and delete all copi=
es
> of the message.
>

--f403045e1d90211ef60554b068a3
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><div>Hi,<br><br></div><div>unfortunately, it is not p=
ossible to convert a DataStream into a DataSet.<br></div>Flink&#39;s DataSe=
t and DataStream APIs are distinct APIs that cannot be used together. <br><=
/div><div><br>The FlinkML library is only available for the DataSet API. <b=
r>There is some ongoing work to add a machine learning library for streamin=
g use cases as well, but this is still in an early stage and mostly focusin=
g on model serving on streams, i.e, applying an externally trained model on=
 streaming data.<br><br></div><div>Best, Fabian<br></div><br></div><div cla=
ss=3D"gmail_extra"><br><div class=3D"gmail_quote">2017-07-19 19:07 GMT+02:0=
0 Branham, Jeremy [IT] <span dir=3D"ltr">&lt;<a href=3D"mailto:Jeremy.D.Bra=
nham@sprint.com" target=3D"_blank">Jeremy.D.Branham@sprint.com</a>&gt;</spa=
n>:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-=
left:1px #ccc solid;padding-left:1ex">


<div link=3D"#0563C1" vlink=3D"#954F72" lang=3D"EN-US">
<div class=3D"m_3758159387246532429WordSection1">
<p class=3D"MsoNormal">Hello =E2=80=93<u></u><u></u></p>
<p class=3D"MsoNormal">I=E2=80=99ve been successful working with Flink in J=
ava, but have some trouble trying to leverage the ML library, specifically =
with KNN.<u></u><u></u></p>
<p class=3D"MsoNormal">From my understanding, this is easier in Scala [1] s=
o I=E2=80=99ve been converting my code.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">One issue I=E2=80=99ve encountered is =E2=80=93 How =
do I get a DataSet[Vector] from a DataStream[MyClass]?<u></u><u></u></p>
<p class=3D"MsoNormal">I=E2=80=99ve attempted to use windowing, but scala i=
s completely new to me and I may need a push in the right direction.<u></u>=
<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">The below code executes properly, I=E2=80=99m just u=
nsure of the next step.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><img id=3D"m_3758159387246532429Picture_x0020_1" src=
=3D"cid:image001.png@01D30086.8C42C4C0" height=3D"215" width=3D"769"><u></u=
><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">I=E2=80=99ve also seen an example [2] that looks lik=
e something I need to implement =E2=80=93 especially the PartialModelBuilde=
r.<u></u><u></u></p>
<p class=3D"MsoNormal">Am I on the right track? <u></u><u></u></p>
<p class=3D"MsoNormal">Thoughts?<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">Thanks!<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">[1] - <a href=3D"https://stackoverflow.com/questions=
/44039857/is-there-a-apache-flink-machine-learning-tutorial-in-java-languag=
e/44040819#44040819" target=3D"_blank">
https://stackoverflow.com/<wbr>questions/44039857/is-there-a-<wbr>apache-fl=
ink-machine-learning-<wbr>tutorial-in-java-language/<wbr>44040819#44040819<=
/a><u></u><u></u></p>
<p class=3D"MsoNormal">[2] - <a href=3D"https://github.com/apache/flink/blo=
b/master/flink-examples/flink-examples-streaming/src/main/scala/org/apache/=
flink/streaming/scala/examples/ml/IncrementalLearningSkeleton.scala" target=
=3D"_blank">
https://github.com/apache/<wbr>flink/blob/master/flink-<wbr>examples/flink-=
examples-<wbr>streaming/src/main/scala/org/<wbr>apache/flink/streaming/scal=
a/<wbr>examples/ml/<wbr>IncrementalLearningSkeleton.<wbr>scala</a><u></u><u=
></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Sprint Sans Medium&=
quot;,sans-serif;color:#404040">Jeremy D. Branham
</span><span style=3D"font-family:&quot;Sprint Sans Medium&quot;,sans-serif=
;color:#404040"><u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:8.0pt;font-family:&quot;Spr=
int Sans Medium&quot;,sans-serif;color:#404040">Technology Architect - Spri=
nt
<br>
O: <a href=3D"tel:(972)%20405-2970" value=3D"+19724052970" target=3D"_blank=
">+1 (972) 405-2970</a> | M: <a href=3D"tel:(817)%20791-1627" value=3D"+181=
77911627" target=3D"_blank">+1 (817) 791-1627</a></span><span style=3D"font=
-size:8.0pt;color:#404040"><u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:8.0pt;font-family:&quot;Spr=
int Sans Medium&quot;,sans-serif;color:#404040"><a href=3D"mailto:Jeremy.D.=
Branham@Sprint.com" target=3D"_blank"><span style=3D"color:#404040">Jeremy.=
D.Branham@Sprint.com</span></a>
</span><span style=3D"font-size:8.0pt;color:#404040"><u></u><u></u></span><=
/p>
<p class=3D"MsoNormal"><span style=3D"font-size:8.0pt;font-family:&quot;Spr=
int Sans Medium&quot;,sans-serif;color:#404040">#gettingbettereveryday</spa=
n><span style=3D"font-size:8.0pt;color:#404040"><u></u><u></u></span></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
<br>
<hr>
<font size=3D"1" color=3D"Gray" face=3D"Arial"><br>
This e-mail may contain Sprint proprietary information intended for the sol=
e use of the recipient(s). Any use by others is prohibited. If you are not =
the intended recipient, please contact the sender and delete all copies of =
the message.<br>
</font>
</div>

</blockquote></div><br></div>

--f403045e1d90211ef60554b068a3--

--f403045e1d90211ef70554b068a4
Content-Type: image/png; name="image001.png"
Content-Disposition: inline; filename="image001.png"
Content-Transfer-Encoding: base64
Content-ID: <image001.png@01D30086.8C42C4C0>
X-Attachment-Id: 75c289b236ab9988_0.1
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--f403045e1d90211ef70554b068a4--