Multi Output from single Row Filter Node

Hi All,

I's Sathish and new to Knime.

We are using Knime for Text Processing.

Now, we have a data in a document which contains data like A,B,C,D as key words along with other some text.

For Ex. data we have like as follows...

This is text data of A id 890
This is text data of A id 890
This is text data of A id 891
This is text data of A id 892
This is text data of B id 887
This is text data of B id 800
This is text data of B id 876
This is text data of C id 844
This is text data of A id 832

For this we used Row Filter Node and using regex we filtered" A" rows and again we used Row Filter pointing the source node and filtered for "B", but returned empty data set. the problem is , the second filter searching from result set of A data instead of Source data..

How to solve this issue? Whether the ongoing approach is correct or need to do some other approch to achieve the same?

Please guide me with your inputs. Looking for the solution.

 

WR,

Sathish R

 

 

 

 

 

 

Hi Satish,

as far as I understand you want to filter a data set with two filters. 1. using *A* as pattern and second using *B* as pattern. You don't need to apply the filter nodes one after the other. You can also connect both to the source node and have them as parallel branches / nodes.

Cheers, Kilian

Thanks Kilian.

Another query Kilian regarding executing Knime workflow in linux command.

I'm having couple of queries in that.

1. How to execute the workflow in Linux using command line? Seen in forum as to use the Batch Mode but still i didn't got any final solution in that. Please guide me for the same.

2. In my workflow, i'm having CSV file as input, if i'm running the workflow from command line in Linux, then how to give the input file ?

Kindly provide me the solutions for the above queries kilian.

Thanks in advance.

WR,

Sathish.R

 

 

 

 

sathhishcse;

If you want to filter just A and B fron the data set, I think there you could use teh node  " Rule-based Row Filter".   It selects just the rows with A and B

In your example should be like this:

$your Colum nname$ LIKE "*This is text data of A id*" => TRUE

$your Colum nname$ LIKE "*This is text data of B id*" => TRUE

 

And select the radiobotton Inclide TRUE matches.

I hope this korks for you.

 

Karelman

 

Hi sathhishcse,

regarding your second questions:

1. you can run KNIME from command line, for details see: https://tech.knime.org/faq#q12

2. to control settings, such as file path you need to use flow variables. These variables can be controlled from the command line call. For details see:

https://tech.knime.org/faq#q12

https://tech.knime.org/forum/knime-general/how-to-pass-a-file-of-workflow-variable-to-a-knime-workflows-in-batch-mode

or

https://tech.knime.org/forum/knime-developers/setting-flow-variables-on-command-line

Cheers, Kilian