I usually use file system to manage data for my deep learning model, but one of my boss told me to make nosql database to manage data.
Datasets I use have m rows, and n columns of count matrix and column, and row names (over 30000 rows, over 30000 columns for each dataset)
I haven't used nosql database unfortunately. So I tried mongodb to make data model that can save column names, and gene names, and count matrix in one document, which causes size limit error.
What kind of approach should I take? Or is there any tips that you manage your data rather than through nosql?