Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
Hi Dataiku Community!
I am a new user and would like to seek your help if there is a specific recipe I can use to copy specific columns from a dataset? The table I am currently working on consists of 200++ fields, an I do not need them all to proceed with the project. This is also causing slowneed when I perform sync and filter, I believe.
Hope to receive your feedback soon, thanks in advance.
Hi, the most common recipe, the Prepare recipe, allows you to easily delete unwanted columns, reorder them and also rename them, among a lot of other transformations. If you switch to the columns view (next to Display, top right) you can select all columns by clicking on the top check box and then click on Action to delete them all. And then add the ones you want manually by deleting the relevant delete step. This will be quicker than deleting one by one if you just want a few columns.
Welcome to the Dataiku community. We are glad to have you join us.
I would agree with @Turribeach that the Prepare recipe is very helpful in trimming columns.
When it comes to overall performance, I would also suggest that you look at the type of storage you are using for your data. Although the built in file based storage is quick and easy to start out. I have found moving to a database to be helpful with scaling. Even on the smallest Dataiku instances I have set up, I tend to use a PostgreSQL server. I have also found that the amount of RAM memory you have is helpful. More is better.