Fig 1.
Diagram for the epidemiological classes in epiABM.
Susceptible (S), Exposed (E), Mild infected (IM), Severe Infected (IS), Hospitalized (H), Recovered (R) and Deceased (D).
Table 1.
Default parameters for epiABM.
Fig 2.
Ensemble based DA for ABMs.
Fig 3.
Total number of agents in different epidemiological categories summed over every neighborhood.
Grey lines indicate ensemble members and colored solid lines their mean. True values are indicated with dashed lines.
Fig 4.
The estimated values as a function of time are shown in solid blue line. Grey lines indicate ensemble members. The true evolution is drawn with a dashed line.
Fig 5.
Relative amount of infected agents per house size as a function of time.
Ensemble members plotted with gray and their mean with red. The true values are represented with green dashed lines. Dotted black lines indicate the distribution of agents in different house sizes regardless of their disease status.
Fig 6.
Evolution of matching metrics over time.
The evolution of the id_matches, housetype_matches, house_matches and housetype_loc_matches to the true state are shown from left to right and top to bottom panels. For every panel, the gray lines represent the metrics for the control simulations and their mean is drawn with a black line. Red (blue) transparent lines correspond to EnKF estimation with the randomized (cascade) agent adjustment method. Semitransparent lines represent the ensemble members while their mean is drawn with a solid line of the corresponding color.
Fig 7.
Diagram for epiABM with categories to account for unreported cases.
Fig 8.
Estimations of active reported mild cases, unreported cases and positivity.
Top panel: evolution of reported active mild cases. Orange solid line is the ensemble mean. Middle panel: Unreported active cases. Teal solid line is the ensemble mean. Bottom panel: Positivity (percentage of agents that would test positive). Red solid line is the ensemble mean. Dots correspond to the testing data generated with the true run. For every panel, gray lines indicate ensemble members and dashed lines the true value of the corresponding variable.
Fig 9.
Estimations of the asymptomatic rate.
The mean of estimation of qA as a function of time is shown with a solid blue line, ensemble members with gray lines, and the true parameter value with a black dashed line.
Fig 10.
Estimations of λ for different housing scenarios.
The mean λ estimates for the different housing distributions are shown with solid colored lines. Ensemble member estimates represented with gray lines. The solid red line indicates the KL divergence between a Poisson distribution and the true geometric distribution used. The dotted red line indicates the minimum of the KL curve.
Fig 11.
The estimated values are shown with a solid blue line and ensemble members with gray lines. The red line corresponds to the 7-day rolling average of daily confirmed cases.
Fig 12.
Total number of agents for different epidemiological categories.
Grey lines indicate ensemble members and colored solid lines their mean.
Fig 13.
Estimated daily new cases for the whole city are shown with orange lines and the corresponding ensemble with gray. Blue dots correspond to the daily confirmed cases.
Fig 14.
Each plot represents one of the 15 CABA communes. Estimated daily new cases are plotted with orange lines and the corresponding ensemble with gray lines. Blue dots correspond to the daily confirmed cases.