RapidMiner 9.3 BETA - Download link and Feedback

sgenzer

Hi all - 

Yep RapidMiner 9.3 Beta is OUT (Studio and Server) and we would love you to download + give us feedback prior to the full release:

LINK TO RAPIDMINER 9.3 BETA DOWNLOAD AND FEEDBACK PAGE

Feel free to chime in on this thread as well if you want to chat. Lots of great stuff to explore!

Scott

varunm1

Hello,

Did anyone try auto model in 9.3? If so, your inputs are much appreciated. I don't know why I am getting an error in the results section. I submitted my log and screenshots in the feedback link.

Thank you

sgenzer

@varunm1 I got the same thing. Submitted a ticket internally. Stay tuned.

[that's why they call it a beta ]

varunm1

Thanks @sgenzer cannot wait to test the new additions in auto model

jczogalla

Just a quick heads-up that we missed a bug in the Azure Blob Storage operators. So if that does not work for you right now, we know the problem and it will be fixed down the line.

sgenzer

@jczogalla can you confirm that the Auto Model problems we're having is also a bug?

jczogalla

Yes, it's a bug and we are already tracking it

[Deleted User]

Hi dear RM friends
If I install this version of RM is the previous version that I have it on my lap top change? or I will try the second type of RM on my system?
( I mean that in this situation I will have two RM or not?)

varunm1

Hellk @mbs,

you are downloading a beta version which doesn't need installation. You can just open the downloaded folder and access RM and you will still have your previous version 

sgenzer

yes just to add on to what @varunm1 says...you can have several versions of RapidMiner co-existing on your machine if you like and go back-and-forth. We of course don't recommend using older versions of the software (for obvious reasons) but sometimes you need to. I have twelve RapidMiner Studio's on my laptop right now including the 9.3.0 beta. That's the life of a manager for a software support community

Some other pieces of advice along these lines:

• I don't recommend running more than one RM Studio JVM at a time on your machine unless you have the horsepower to do this. A standard laptop is just going to bog down.
• If you are going back-and-forth with really older versions, be aware that things are going to break and you're going to get annoyed.

One method that I have developed over time is to keep different .RapidMiner folders on my laptop for different major releases. I have one for RM 7.x, one for RM 8.x, and another for RM 9.x. This helps decrease the number of things that break. When I want to run 7.x, I put all my current .RapidMiner contents of 9.x into a folder called "dot rapidminer 9" and put all the contents of my 'dot rapidminer 7" into the .RapidMiner folder. Then boot. When I'm done, I swap back.

Again none of this is "official" RapidMiner documentation but rather tricks-of-the-trade that I have come up with over the years. No warranty if things get messed up!

Scott

[Deleted User]

sgenzer 12 RM!!!!!!!!!!!! poor your lap top
you are right. Im thinking that if I install any other version what will happen on my project in the previous one
any way sgenzer and varunm1 thank you for your recommendation

IngoRM

Just FYI: the Auto Model has been identified and fixed.  Thanks again for bringing it up.

Enjoy your weekend,
Ingo

User222697

Hi,

a bug here, in the final step of 'automodel', select "Automatic Feature Selection" additional time to "120" and "Automatic Feature Generation" to "High".

Now, select the "Naive Bytes" model and click on RUN. (the issue it can also be produced in other models)

Expected result: The model must work for 120 minutes to generate the most precise attributes.

Actual result: The model gives the results before the expected time, usually in less than 30 minutes.

IngoRM

Hi,

Thanks for the feedback.  In fact, the time limit in this setting is a hard time limit cutting off the computation in case it is hit.  But it is not the only way to stop the feature engineering process.  If the optimization runs for a certain amount of time without any progress whatsoever, it can also stop for that reason for example.  We have changed the wording to better reflect the nature of the time limit and it now reads "Additional Minutes (Maximum)" to better transport the idea that this is the maximum of the additional time but can be sometimes less.  Hope this makes this clearer now.

Thanks again,
Ingo

yousef

I am trying to connect to RM9.3 Jupyter NB

import rapidminer

import os

rm_home="C:\Downloads\rapidminer-studio-9.3.0-BETA2-win64"

connector = rapidminer.Studio(rm_home, rm_stdout=None)

No problems until I ran:

df = connector.read_resource("//Samples/data/Iris")

print("The result is a pandas DataFrame:")

print(df.head())

I always get this error:

FileNotFoundError: [WinError 2] The system cannot find the file specified

IngoRM

Just for your interest: we have published an updated version of the Beta (see link above) if you want to give it another try.  Lots of things have been fixed in this version already, including the Auto Model problems discussed above.

Many thanks for all your support,
Ingo

ershadalibaig_m

LINK TO RAPIDMINER 9.3 BETA DOWNLOAD AND FEEDBACK PAGE - 

i am not able to download the beta product from the above link. It says access denied. Can someone help me on this pls?