Question

Populate entire column based on specific value in the same column

  • 2 April 2018
  • 4 replies
  • 66 views
D

Is there an easy way to populate an entire column (I have multiple columns) based on the value in that column that has a comma in it?

Basically, if I have a column named pole and all attributes in that column have the value of "pole_status" except for one value that has this value: "Con, PLA, MRC, MRE". How can I populate the entire column with "Con, PLA, MRC, MRE".

I've tried Attribute Manager using a Conditional Value to filter out all records that don't have comma in it but then I'm not sure how I can actually use the "Con, PLA, MRC, MRE" value as the output value. Maybe there's a better option. Custom Python might do the trick - but I want to exhaust all of my transformer options before I go that route. Note: I cant simply write "Con, PLA, MRC, MRE" in the AttributeManager since these values will change and there's multiple columns - si it wouldn't be sustainable

I also know I can probably accomplish this by using a multiple FeatureMergers but I definitely don't want to do that (if I dont have to) due to performance (there would probably be 20 + FeatureMergers).

Thanks!

4 replies

lars_de_vries
Badge +3

I think that I would use a TestFilter to find a feature that matches "pole" Contains ','. Subsequently I would use a Sampler to make sure only a single feature will be output through a FeatureMerger (or FeatureJoiner in FME2018). I would set the FeatureMarger/-Joiner to match 1 - to - 1 (making it an unconditional merge) and make all features from "pole" pass. You could use a AttributeRemover to remove the "pole" attribute from the supplier features.

I hope this helps you.

fmelizard
Rank
Userlevel 4
Badge +13

So it sounds like you need to scan all the values for a column in all the rows, and if any of them had a , choose the first one that has the , and use that for as the value. Else leave them as they were. I'd suggest a workflow involving a tester, followed by a VariableSetter when you find the match (i.e. set a global variable named after the column to have the value of the , value). Do this tester/variablesetter for each column you want to do this to. Then route everything to FeatureHolder. After the feature holder use a Variable Retriever to pull out the value for the column and set it (you'll need one for each column). Lastly, you may need to set a default value for each Variable using a Creator followed by VariableSetters if you aren't guaranteed of one row with a ,

 

 

Great scenario. Let us know if this works (and if anyone has a better idea, do share).
D

So it sounds like you need to scan all the values for a column in all the rows, and if any of them had a , choose the first one that has the , and use that for as the value. Else leave them as they were. I'd suggest a workflow involving a tester, followed by a VariableSetter when you find the match (i.e. set a global variable named after the column to have the value of the , value). Do this tester/variablesetter for each column you want to do this to. Then route everything to FeatureHolder. After the feature holder use a Variable Retriever to pull out the value for the column and set it (you'll need one for each column). Lastly, you may need to set a default value for each Variable using a Creator followed by VariableSetters if you aren't guaranteed of one row with a ,

 

 

Great scenario. Let us know if this works (and if anyone has a better idea, do share).
@daleatsafe So there's a couple of ways to accomplish this and your solution seems to work faster than mine - at this point in my workflow (it's a huge workflow) I just wanted to optimize this part. Good enough thanks!

 

D

I think that I would use a TestFilter to find a feature that matches "pole" Contains ','. Subsequently I would use a Sampler to make sure only a single feature will be output through a FeatureMerger (or FeatureJoiner in FME2018). I would set the FeatureMarger/-Joiner to match 1 - to - 1 (making it an unconditional merge) and make all features from "pole" pass. You could use a AttributeRemover to remove the "pole" attribute from the supplier features.

I hope this helps you.

 

@lars_de_vries this works well but looks like Dale's solution was a bit faster - faster than mine as well

Reply


Community Stats

28,662
Posts
109,326
Replies
36,528
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.