balance of num_partners

[andrew-phillips-1]

I was looking at the rred_balance code. I've played around with certain changes with the intent that num_partners balances more by gender (albeit with a lot of period to period variability). This is in sexual_behaviour.py in my branch andrew__jan24. It seems like it is taking two time steps for changes to rred_balance to properly take effect in reversing any gender imbalance which I assume is in some way due to the order of the operations.