I am working with a dozen of CSV files containing financial simulations in form of 100'000x100 matrices. Each file is around 100Mb large.
Until now, only analyse these files with R (or Matlab).
Since I only perform elementary operations on these matrices, I am thinking about importing them in a database and perform the same calculations with some SQL queries.
Is it a bad practice to use a database server (SQL Server in my case) to perform such mathematical computations (conditional sum, product,...) ?
The server is slower than a program in R, but I find it less prone to error, specially when I have a lot of CSV files.