brunet | a P2P library developed at the University of Florida

I've read the others related questions but mine is unique because of it's structure.

My App is storing about 10,000+ users with profiles defined by many parameters (gender, weight, height, hair color, eye color, dancing skills... and so on, about 100 attributes, lets say).

The application construct a Filter Form with these attributes. User is filtering database using this form so constructs a Query with many sub-queries, one for each filter used.

The problem is that using more than 8-9 filters, the engine will crash into a very long response (I had to kill the process after 30m of waiting).

So, this is the structure of Database

Table def_attributes (here are the attributes definitions)

• id ---> uses as attr_id in values table

Table utilizatori (users definitions, only column activ is used now)

• id ---> is named user_id in rest of tables
• activ ---> is 1 if user is active and will be displayed (column indexed)

Table val_atribute (storing the values of attributes for each user)

• attr_id ---> the attrID of a filter (column indexed)
• attr_value ---> value of the attribute
• user_id (column indexed)

For example, here is a query constructed by the Filtering form, that lags out:

I am trying to make a Python interface for the brilliant NMF package - https://cran.r-project.org/web/packages/NMF/NMF.pdf (as it is much more flexible than the Python options). So far so good.

I come up with something like this:

My code does read and write the file, but it is not on a new line for every value and instead prints every value in one line.