MGSurvE 0.1.1.1

MGSurvE

MGSurvE: Mosquito Gene SurveillancE

MGSurvE is a project oriented towards the optimization of traps' placement in complex heterogeneous landscapes in an effort to minimize the time to detection of genetic variants of interest.

Under Construction :construction: Please check back in a couple of weeks!

Please have a look at the documentation for more info and our pypi package for installation and versions.

Authors and Funders

• Lead and Dev: Héctor M. Sánchez C.
• Active Devs: Elijah Bartolome
• PIs: David L. Smith, John M. Marshall

       

---

Progress and Features

• Setup pkg skeleton
  • Setup pypi pkg
  • Setup CI
  • Setup autodoc
• Landscape generation and variables update
  • Basic landscape and matrices init
  • Auto-calculate distance matrix
  • Auto-calculate migration matrix
  • Landscape onbjects constructors
  • Auto distance and migration updates on change
  • Update traps migrations in place (avoids memory shifts)
  • Update traps migrations with new array if traps number changes
• Plots
  • Landscape
    • Auto-assign markers with point-type
    • Auto-assign colors with trap-type
    • Auto-assign edge to trap if non-movable
    • Auto-plot trap radii
    • Spherical coordinates projection
    • Directed networks
  • Migration matrices
  • Block Migration matrices
  • Remove frames and axes
• Genetic Algorithm
  • Init chromosome
  • Fitness function
    • Canonical form
    • Get steady absorbing states
    • Put the cost function together
    • Markov's fundamental matrix with no re-ordering
    • Test different fitness functions
  • Code new mutation functions
  • Code new crossover functions
  • Implement full optimization loop
• Code efficiency
  • Process traps movement in place
• Tests
  • Points and traps numbers are updated correctly
  • Check matrices for Markov validity
  • Test landscape with external migrations
  • Changing number of traps doesn't affect migration part
  • Selective mutation
• Wishlist
  • Non-movable traps
    • Add to object properties
    • Add to mutation
    • Add to crossover
  • Different trap types
  • Male/Female kernel
  • Point-processes generation
  • Parallelize fitness function evaluation