licensing question...
hi,
1.I got the educational license, but whats different to the free license? It just looks as its completely the same?
2. is it somehow possible to get a dicsount for the commercial versions if you are a student? or for universities? I would like to 4 cores version, but 5000 € is a bit too much for me, any chance to get discount? (maybe my university will purchase a license with better conditions?)
Best Answers
-
BalazsBarany Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert Posts: 955 Unicorn
Being able to use more cores won't dramatically accelerate your processes, as many operations are not completely parallelized.
If your process takes days, it will also take days if executed on 4 CPUs.
I consider the non-commercial license a great gift to the academic and research community.
If your time is measured in money, as it is in companies, the commercial product gets more interesting.
0 -
IngoRM Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert, Community Manager, RMResearcher, Member, University Professor Posts: 1,751 RM Founder
Hi Fred,
There are NO differences between the free and the commercial versions in RapidMiner v7.2. The free version comes with all features and functions also offered by the paid version. So if you saw some functionality in some documentation it might have been moved somewhere else, been replaced, or even removed completely from the product. It would be helpful if you could point out where you found this information so we can update this if possible.
But there are two limitatations for the free version:
- The free version of RapidMiner is limited to work on a maximum of 10,000 data rows (as input for operations but also for viewing them). If you are part of the educational program of RapidMiner, the number of rows is not limited.
- The free version of RapidMiner can only use 1 logical processor for computation (if you for example have a 2-core machine with hyperthreading this leads to 4 logical processors on your machine).
The first thing is pretty obvious: the more data you analyze the better your models are. The second thing means: the more logical processors you are using, the faster the computation will be done if things are parallelized. For modeling schemes like Decision Trees, Generalized Linear Models, or Gradient Boosted Trees among others, you can see a speed-up of almost the number of logical processors you have (in our example above: almost 4X). In other cases, e.g. for deep learning, we use the extra power to make more computations in the same time. So you won't see a dramatic speedup but will get better models. Future versions of RapidMiner will have even more parallelized methods and will even offer to compute complete processes in the background as well.
Hope this helps,
Ingo
1
Answers
if you are student or academci professional, check this out,
You can work with Rapidminer team to verify your credentials etc
https://rapidminer.com/academia/educational-program/
do you read what people are writing ? ;P
as I said, I already have the academia license, but it seems to have no benefit at all, except maybe a little bit more collaboration in teams..
my question was furthermore, if there are any better conditions for students for the commercial licenses..
btw... the order of postings now seems to be confusing ;D later ones appear on top.. hm this forum gets weirder every day
Question back to you Fred, did you read the link :P
Jokes apart, but to summarize..
The academia license from 7.2 release provides full product unlimited access, without any restrictions, if you can satisfy the requirements.
Including unlimited studio, server, radoop..
So its everything for free, but only community support and certain requirements about references, etc
Details are here
For your pricing request, I had already reached out to my sales team, so someone will reach out to see if there is a possibility
hi,
if I can read correctly, I see it is limited to 1 (one) logical processor only, I don't know if its possible to do things in parallel like that, of course its limited therefore
I would like more processors as my processes with opt. parameters take sometimes days to weeks... that would speed things up.
if it was all the same as with commercial licenses, nobody would need them
Ok, let me take my words back.. memory unlimited but limited to 1logical core for processing power
I just confirmation from my internal team and someone will reach out to you to see how we can get around your concerns.
ok I accept that as a solution, but 1 more question: why is the parallel extension in Rapidminer 5 not available for Rapidminer 7 ? does that mean version 5 can do stuff in parallel, but not 7? and therefore, version 5 is still faster than 7?
Internal structures of RapidMiner Studio changed between version 5 and 7. Also, the extension did quite low level stuff and wasn't very reliable. It also didn't work very well (never was able to fully use all 4 cores of my machine).
That extension was right for the 5 era but it's not applicable nowadays. More and more machine learning algorithms are updated to parallelized versions (RapidMiner updated Decision Tree a few versions ago, for example), so the need for a separate extension is disappearing.
ok so but then the commercial versions, do they have any effect regarding parallelism and 4 cores? or is it just a marketing gag ?
I was thinking hard to get a discounted license, speaking of that, I wanted to inform myself of the options for students / academics / universities licenses, is there a discount possible? I will probably not pay it myself, my university will.. and on how many machines am I allowed to use the license on?
Sort order now restored @Fred12
Apologies for any weirdness!
no problem, still waiting for an answer.. besides, the "print process to pdf" and "pause process" button, are only available in commercial versions? I seen the description in studio documentation, but cannot see those in Version 7.2 ?
ok and what is the difference between the more-core versions of 7.2 and the "parallelize operators" extension in Rapidminer 5? does it have the same effect?