Error in cv.make_synthpop

[cmerenstein]

Hello,

I think there's an error in the newest version cv.make_synthpop, or at least a change in behavior that I can't find documentation on.

I made a minimal example of the issue in this colab: https://colab.research.google.com/drive/1fG8MxWYqhv44wpEDg0QzRpL1ySysfSp6?usp=sharing
which gets an error when running on the newest version of covasim, but not when installed with version 3.0.7. Below is more information on the bug:

I am trying to use make_synthpop to add a population to a Sim object, with a synthpops population dict as my reference. I had been using this in a script previously, but with the latest update it no longer works, giving me an error originating in the "make_random_contacts" function.

The error message is:

TypeError                                 Traceback (most recent call last)

<ipython-input-6-465ff658c7d6> in <module>()
      3                         use_waning=True))
      4 
----> 5 sim_household.popdict = cv.make_synthpop(sim=sim_household,population=sc.dcp(pop_dict))

1 frames

/usr/local/lib/python3.7/dist-packages/covasim/population.py in make_random_contacts(pop_size, n, overshoot, dispersion, mapping)
    223 
    224     # Precalculate contacts
--> 225     n_all_contacts  = int(pop_size*n*overshoot) # The overshoot is used so we won't run out of contacts if the Poisson draws happen to be higher than the expected value
    226     all_contacts    = cvu.choose_r(max_n=pop_size, n=n_all_contacts) # Choose people at random
    227     if dispersion is None:

TypeError: unsupported operand type(s) for *: 'int' and 'dict'

[cliffckerr]

Well spotted @cmerenstein , we'll fix this in the 3.1.2 release later this week. Thanks!

[cliffckerr]

Thanks for your patience @cmerenstein and apologies it took much longer than we hoped to fix! We ended up completely revamping the way population loading/saving is handled, but the new system should be much easier to use. You can now do things like:

import covasim as cv
import synthpops as sp

n = 2e3
people = sp.Pop(n=n)
sim = cv.Sim(pop_size=n, pop_type='synthpops', people=people).run()
sim.people.plot()
sim.plot()

Full changes listed here.