Skip to contents

Assimilation of data (Bayesian) using optimal statistical inference, a modified Kalman Filter.

Usage

assimilateData(
  layers,
  linearInterpolationMatrix,
  prevalenceData,
  healthZoneCoordinates,
  psi.diagonal,
  QHt,
  HQHt
)

Arguments

layers

The SpatRaster object with layers Susceptible, Vaccinated, Exposed, Infected, Recovered, and Dead.

linearInterpolationMatrix

a linear forward interpolation operator matrix with as many columns as cells in the SpatRaster layers, and as many rows as health zones for which coordinates are provided (when creating a matrix for use with one state vector). The matrix is either a trivial matrix as described elsewhere (see linearInterpolationOperator), or two partitions in a sparse, block diagonal matrix.

prevalenceData

A "situation report" dataframe. The first column provides the date of the officially reported, observed incidence of the disease, in ISO format (YYYY-MM-DD). MAYBE TODO: enforce the startDate parameter to be one week prior to the first observed data?


    Date    Beni  Biena  Butembo  Goma  Kalunguta
    05-Aug  3     0      2        0     0
    12-Aug  2     0      0        0     0
    20-Aug  1     0      0        0     0
    26-Aug  5     0      0        0     0
    02-Sep  8     0      0        0     1
    09-Sep  5     0      2        0     0
    16-Sep  5     0      3        0     0
    23-Sep  4     0      1        0     0
    02-Oct  10    0      1        0     0
    07-Oct  14    0      4        0     0
    15-Oct  28    0      2        0     1
    21-Oct  19    0      3        0     0
    28-Oct  28    0      8        0     0
    04-Nov  16    0      6        0     1
    11-Nov  14    0      0        0     21
  
healthZoneCoordinates

a table of values giving the latitude and longitude coordinates for health zones in the country of interest. See the description of the healthZoneCoordinates argument in the function SVEIRD.BayesianDataAssimilation for more details.


    HealthZone    Latitude    Longitude
    Alimbongo     -0.365515   29.1911818
    Beni          0.49113     29.47306
    Biena         0.57923     29.115633
    Butembo       0.140692    29.335014
    Goma          -1.658271   29.220132
    Kalunguta     0.323085    29.354677
    Katwa         0.116985    29.411838
    Kayna         -0.603936   29.174066
    Kyondo        -0.005622   29.408813
    Lubero        -0.15633    29.24057
    Mabalako      0.461257    29.210687
    Manguredjipa  0.353433    28.765479
    Masereka      -0.133333   29.333333
    Musienene     0.04022     29.26246
    Mutwanga      0.32893     29.74576
    Nyiragongo    -1.516667   29.25
    Oicha         0.698681    29.518834
    Pinga         -0.9830504  28.687911
    Vuhovi        0.1416      29.4075
    Ariwara       3.136873    30.706615
    Bunia         1.566667    30.25
    Komanda       1.367496    29.774322
    Lolwa         1.352969    29.498455
    Mambasa       1.359731    29.029226
    Mandima       1.35551     29.08173
    Nyakunde      1.431271    30.029283
    Rwampara      1.4053      30.3449
    Tchomia       1.4412      30.4845
  
psi.diagonal

A replacement value for elements of the Psi matrix' diagonal which are zero.

QHt

the tcrossprod of the Q and H matrices.

HQHt

the product of the H matrix and QHt.

Value

a SpatRaster with Infected and Exposed compartment layers.

Author

Bryce Carson