I try to master the Talend data preparation software .
Is it possible to edit cells by :
- By simply writing within ?
- Summing two columns ? (I can concatenate, but i also would like to sum the values wich are inside the columns)
- Making other kind of operations ?
If it's possible, I don't know how to do it with the functions
When I permute two columns, this change is not added to the recipe, isn't it ?
What are the software limits?
Thank you for your attention.
- Editing a cell: sure! just double click on the cell. Hit enter to commit the change or Esc to cancel. Bear in mind everything you do adds a step in the recipe (left panel) so editing a cell is no exception: it simply generates a "replace" function. Also check out the checkbox "apply to all" when you edit in a cell. It will automatically extend the replacement to all cells with the same original value throughout the column.
- Summing too cols or other kind of operations: you have a point that "sum" is not found when you search for functions (in the 'find a function' field, top of the function list). And neither is "operation". We should add those 2 keywords in our search index. The function exists though, it is called: "Add, Multiply, Subtract or Divide".
- Columns order: no it's not captured in the recipe. I kind of agree it should though. Something we probably will look into.
- Software limits: the main limit is explained in my post here https://www.talendforge.org/forum/viewt … p?id=48262
Thank you very much ctoum .
I have an other question :
I added two columns with your function, but the result is that when I do this, I get 3 columns (the two firsts, and a third one which is the "result column" . So I deleted the two firsts columns, and I think that this is not an action which is accepted by the softawre : Now it's impossible to open my recipe.
Is it possible to make an operation and then remove the columns we use ?
Thank you for your attention .
Strange that you can't open the preparation after deleting the columns. This is totally allowed and should work.
Any chance you would share a sample of your data and a reproduceable scenario with us? This is a public forum but you could send us details about this issue directly at: email@example.com