Nested if else condition
I have a scenario where I have create a if else condition in dataiku which is having around 6000 conditions. The scenario is like:
if ( col1 == 'A' , 1 ,if( col1 == 'B',2, if(col1 == 'C',3,0)))
The formula option with if condition gives stackoverflow error. Is it the length of the if else conditions that is causing the error or can it be something else.
Also I was trying find and replace as an alternative option, is there an option in find and replace to enter all these 6000 strings and its replacements from a file instead of manually doing a copy paste of each of these conditions. Also whats the actualuse of 'raw text edit' option
Thanks
Answers
-
tgb417 Dataiku DSS Core Designer, Dataiku DSS & SQL, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Neuron 2020, Neuron, Registered, Dataiku Frontrunner Awards 2021 Finalist, Neuron 2021, Neuron 2022, Frontrunner 2022 Finalist, Frontrunner 2022 Winner, Dataiku Frontrunner Awards 2021 Participant, Frontrunner 2022 Participant, Neuron 2023 Posts: 1,601 Neuron@anthony
,Welcome back to the Dataiku Community.
Yes, it is likely that 6000 nested if-else statements, will cause you a problem. 6-10 nested ifs I've done but not 6000. (It would be interesting to know from staff what the "()" Nesting limit of the shaker language used for Visual Recipies might be.)
Another way that may be helpful to achieve the goals you appear to be going for is to use a join recipe.
In your example above, I see that the column you are evaluating is being held steady on "col1". If that is your actual intent and not just a simplified example.
Then, you might create a manually edited table in Dataiku OR import an Excel Spreadsheet in a two column format seen below:
Col1_Value New_Column_Value A 1 B 2 C 3 ... ... (The first column contains the value you are looking for in your table. The second with the value you wish to put in the new table column.)
This table would extend to your 6000 entries or way further.
You could use a join visual recipe to connect this "business logic" table to your original table. When the value in the "Col1_Value" is the same as the value in the "Col1" or your original table. The join recipe would connect the two records and add the value as a new column in the resulting table. This is sort of like VLookup that you might be accustomed to in MS Excel. (But more powerful in some ways.)
Hope that helps a bit.