I tried sorting in pipeline builder a particular dataset, but when the object gets indexed the object’s order is completely different than the backing dataset’s. I initially thought this is a random shuffle but when I present the object in a table in workshop, there is a consistent order (unsorted) whenever I refresh the page.
This brings up the question - in Palantir’s indexing process in an object set, how is the order determined? Is it completely random? Is there any way to preserve sort in pipeline builder to translate to the object level? (I feel like this defeats the purpose of sorting in pipeline builder)…
What is the question behind the question? What are you trying to achieve with a sorted object? If you need to sort an object table in workshop, there is a setting for this in the table’s configuration pane (see below). If not, can you please share what your unique use case is that uses sorted objects?
Thanks for your response - I am trying to export an object set of large size in excel format from workshop - like 100k objects / rows and want it to be sorted.
It appears that Workshop doesn’t allow exporting an object set of that size after it has been sorted due to capacity or memory limitations.
