Pages: [1] 2 3 ... 10
 1 
 on: July 02, 2015, 06:13:14 PM 
Started by trailblazer - Last post by trailblazer
Hi guys! I can't seem to find any good documentation on the python module for rapid miner. How are we supposed to format return values? How do I see my print statements?

Thanks in advance!

 2 
 on: July 02, 2015, 04:47:29 PM 
Started by eisioriginal - Last post by eisioriginal
Hi,

i'm currently implementing matrix operations within R. I'm passing large high dimensional matrices to the R script which i want to convert to a sparse matrix within R. Therfor i use the "Matrix" package. Once i try some simple operations with such a matrix as x > 0 or Matrix multiplication i get some eroors. I checked the RapidMiner created object by storing it as a file with the "save" operation in my R script within Rapid Miner. When i import this object to a ragular R environment i can carry out all operations on a sparse matrix based on this object.

Is there any issue with external packages, especiially sparse matrices? Can i do matrix operations in RapidMiner as well?

All the best

Andreas

 3 
 on: July 02, 2015, 07:43:54 AM 
Started by sukh - Last post by sukh
hi
i want to know if i have a text files having hindi words(indian language) for replacement.
i have written the same text file  for English in works but for hindi it gave me an error.
Kindly let me know if it could be possible to do stemming of an hindi word.

Regards:
 Sukh

 4 
 on: July 02, 2015, 06:05:33 AM 
Started by sukh - Last post by sukh
Actually if the label is negative then why the magnitude of sentiment comes in positive and vice versa. I could not figure out this.
 Regards:
 Sukh

 5 
 on: July 02, 2015, 02:58:32 AM 
Started by Geoffb - Last post by Geoffb
I think it would be useful to have more guidelines to help users decide which models/clustering methods etc might be applicable for their particular data and also the pre processing methods required.  Could be in a wizard type questionnaire and/or depicted in a decision tree? For instance the data might have lots of nulls but I am not sure how to easily filter our which approaches can still be used?

For instance some key questions might be:

Is your data all numerical, all categorical or combo?
If Categorical data would you like suggestions on how to convert categorical to numerical?
Does your data have null values?
If has nulls do you want to remove, pre-process or only use algorithms that can handle nulls?
Are you wanting to understand your data or predict outcomes?
Do you have training data for supervised methods and modelling?
Do you have lots of information dimension which you need to reduce?
If so then do you want recommendations on how to test your results?
Would you like a suggestion on the approach based on these question??
A typical application could also be depicted such as for Churn prediction?

 6 
 on: July 01, 2015, 08:45:40 PM 
Started by MBA_Data_Miner - Last post by MBA_Data_Miner
question for experts:

if  your dependent variable is something like success percentage  ( calculated on two other variables total attempts and successful attempts). Success percentage is calculated as follows:

IE  Success Percentage (S) = Total successful attempts (A) / Total attempts  (T)


Does it make sense to exclude total successful attempts (A) as an independent variable in predictive models using S as the dependent variable? If so, why?

I am thinking it should be excluded to avoid allowing the model to "cheat" by using a variable that is not completely independent to the dependent variable.

 7 
 on: July 01, 2015, 11:36:20 AM 
Started by w.fahey - Last post by w.fahey
Hi all

I am using a classification model where the performance output is a confusion matrix.

But the number of classified records in the confusion table is much less that what was presented to the model. Would anybody have any idea why this would happen? I can send an example if required?

Thanks

 8 
 on: July 01, 2015, 09:25:06 AM 
Started by Xiaoren - Last post by Marco Boeck
Hi,

you can make the app accessibly via direct link w/o having to log in. See anonymous access in the general settings for an app.

Regards,
Marco

 9 
 on: July 01, 2015, 06:12:35 AM 
Started by Xiaoren - Last post by Xiaoren
Hi,

I am working on designing app in the rapidminer server. I want to let my client see their app(or report) and only their own app(or report) as they log in the rapidminer server?

What I think is first create the app by the administer or any one has the permission to edit report. Then change the permission for the client, so when the client log in, they can view the report. But as I try that, client can't open anything, all the directory is empty.

When I search for it, I found solution for "service", but I didn't found similar solution for app. Can anyone give some suggestions to me or preferably a procedure to do that?

Best,
Xiaoren

 10 
 on: June 30, 2015, 11:54:47 AM 
Started by yongie - Last post by yongie
Hi there,

I have both RapidMiner Server and Studio installed on same machine, if I launch Studio on the machine itself, there is no problem, if I launch Studio from a remote PC through RapidMiner Server web interface, there is no Radoop under Tools and no Hive Management at the tool bar. Any one have same issue?


Pages: [1] 2 3 ... 10