Inheritance utilities

[GuettnerBianca]

Implement analysis and record functions for IBLGCA with family tracking.

• [x] number of families in total and alive
• [x] population per family
• [x] total population
• [x] family inheritance tree (how to represent? - not binary)
• [x] which generation each family belongs to = how many mutations occurred
• [x] population per family generation
• [x] population per initial ancestor
• [x] add filter for population cutoff in each of these functions (neglect families which never exceeded a certain population size)
• [x] propagate family properties down to cell level
• [ ] diversity indices: Shannon index, evenness of Shannon index, richness (number of families with live members), Berger Parker index, Gini Simpson index, Simpson index, Hill numbers of given orders
• [ ] initialize nodes with given number of cells from each family
• [ ] process custom nodes with a given family structure (needs updates to init_nodes and init_families) - take care that the family structure respects the internal LGCA conventions because otherwise interpretation functions crash
• [ ] simulation run stopping condition like "only cells of one family alive", "only cells from one ancestor alive"

Plots:

• [x] Muller plots (plt.stackplot?) : https://phylo-baltic.github.io/baltic-gallery/advanced-muller-plots-raw/
• [ ] mutation pie plot - implementation could use IdentityColourMapper and PropertyColourMapper from lgca.plots
• [ ] family ancestry tree - bubbles for each clone, size according to population size

[GuettnerBianca]

Is it needed to propagate the properties of a family down to the cell level?

[sisyga]

yes

[GuettnerBianca]

• population per family - implemented as lgca.calc_family_pop_alive()
• family inheritance tree - basic implementation via lists, navigation from root: family_props['descendants'], navigation from leaves: family_props['ancestor']

[GuettnerBianca]

removed [ ] record total population, number of families (total and alive), population per family/generation/ancestor incl. after filtering with population cutoff because:

• lgca.timeevo(recordN=True) already records the total population size and preemptive filtering loses information about families that grow large enough later on
• all other records including filtering can be obtained by setting lgca.timeevo(recordfampop=True), which records the family populations over time in the attribute lgca.fam_pop_t, and calling the respective functions (see below) that also have a option to filter by absolute or relative population cutoff.
  • number of families in total: lgca.num_families_total_t()
  • number of families alive: lgca.num_families_alive_t()
  • population per generation: lgca.calc_generation_pop_t()
  • population per ancestor: lgca.calc_init_families_pop_t()

Recording everything separately is a keyword mess and can be added if needed