madlib-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rahul Iyer <ri...@pivotal.io>
Subject Re: MADlib's Random Forest Classification and Regression Example
Date Tue, 19 Jul 2016 20:25:56 GMT
Hi Nurur,

I'm happy it worked out.

I don't have a concrete reason for suspecting a problem with the database.
The error message indicated that the madlib.bytea8 composite type was
missing or invalid. For HAWQ < 2.0, that type comes pre-built in the
database catalog and is not installed with MADlib. This led me to guessing
that the catalog was corrupted, probably due to that type being dropped or
changed by a superuser.

Rahul



On Tue, Jul 19, 2016 at 11:17 AM, Nurur Rahman <nurur_rahman@glic.com>
wrote:

> Hi Rahul,
>
>
>
> I changed the database and was able to run the model successfully. Thanks
> a lot for this useful tips.
>
>
>
> For a sake of completeness and learning, how did you know that a ‘corrupt
> database’ might be a
>
> reason behind? Would you like to share a bit of your experience/thought on
> this.
>
>
>
> Nurur
>
>
>
> *From:* Rahul Iyer [mailto:riyer@pivotal.io]
> *Sent:* Monday, July 18, 2016 7:45 PM
> *To:* user@madlib.incubator.apache.org
> *Cc:* Frank McQuillan <fmcquillan@pivotal.io>
> *Subject:* Re: MADlib's Random Forest Classification and Regression
> Example
>
>
>
> Hello Nurur,
>
>
>
> I tried reproducing your error on an existing HAWQ 1.3.0.0 installation
> with a new database and was unable to do so (see attached log for
> successful query results).
>
>
>
> It's possible something is corrupted in the database you're using. Was the
> madlib schema previously dropped in your database? Would it be possible for
> you to test this again on a fresh database?
>
>
>
> Best,
>
> Rahul
>
>
>
> On Wed, Jul 13, 2016 at 8:07 AM, Nurur Rahman <nurur_rahman@glic.com>
> wrote:
>
> Hello,
>
> I am going through Random Forest example given in MADlib's User
> Documentation:
>
>
> http://madlib.incubator.apache.org/docs/latest/group__grp__random__forest.html#examples
>
>
>
> In both cases when I run “SELECT madlib.forest_train( … … )” exactly as it
> is shown in
>
> steps 1 and 2 in the example page, I get the following error:
>
>
>
> ERROR:  plpy.SPIError: Function
> "madlib._dt_apply(madlib.bytea8,madlib.bytea8,madlib.bytea8,smallint,smallint,smallint,boolean,integer)":
> Invalid type conversion. Simple type supplied but backend expects composite
> type. (plpython.c:4648)
>
> CONTEXT:  Traceback (most recent call last):
>
>   PL/Python function "forest_train", line 39, in <module>
>
>     sample_ratio
>
>   PL/Python function "forest_train", line 553, in forest_train
>
>   PL/Python function "forest_train", line 2192, in _tree_train_using_bins
>
>   PL/Python function "forest_train", line 1162, in _one_step
>
> PL/Python function "forest_train"
>
>
>
>
>
>
>
> Relevant Info:
>
> "PostgreSQL 8.2.15
>
> (Greenplum Database 4.2.0 build 1)
>
> (HAWQ 1.3.0.0 build 12954)
>
>
>
> "MADlib version: 1.9, git revision: rc/v1.9-rc1
>
>
>
>
>
> Can anyone please explain what the reasons behind this error message? Your
> help will be appreciated.
>
>
>
>
>
> Thanks.
>
> Nurur
> ------------------------------
>
> ----------------------------------------- This message, and any
> attachments to it, may contain information that is privileged,
> confidential, and exempt from disclosure under applicable law. If the
> reader of this message is not the intended recipient, you are notified that
> any use, dissemination, distribution, copying, or communication of this
> message is strictly prohibited. If you have received this message in error,
> please notify the sender immediately by return e-mail and delete the
> message and any attachments. Thank you.
>
>
>
>
>
> --
>
> ---------------------------------------------------------
> Rahul Iyer
> Principal software engineer | Predictive Analytics
>
> *Pivotal **A new platform for a new era*
>
> ------------------------------
> ----------------------------------------- This message, and any
> attachments to it, may contain information that is privileged,
> confidential, and exempt from disclosure under applicable law. If the
> reader of this message is not the intended recipient, you are notified that
> any use, dissemination, distribution, copying, or communication of this
> message is strictly prohibited. If you have received this message in error,
> please notify the sender immediately by return e-mail and delete the
> message and any attachments. Thank you.
>
>


-- 

---------------------------------------------------------
Rahul Iyer
Principal software engineer | Predictive Analytics

*Pivotal**A new platform for a new era*

Mime
View raw message