computeDistrParams: Computes the parameters of the population counts...

In bogdanoancea/inference: R package for computing the number of individuals in the target population conditioned on the number of individuals detected by the MNO and auxiliary information.

Description

Computes a series of parameters needed to build the target population counts distribution. For a complete description of these parameters an interested reader can consult the description of the methodological framework https://webgate.ec.europa.eu/fpfis/mwikis/essnetbigdata/images/f/fb/WPI_Deliverable_I3_A_proposed_production_framework_with_mobile_network_data_2020_05_31_draft.pdf.

Usage

computeDistrParams(
  omega,
  popRegFileName,
  pntRateFileName,
  regsFileName = NULL,
  gridFileName = NULL,
  rel_bias = 0,
  cv = 1e-05
)

Arguments

omega	The deduplication factors. They are computed by computeDeduplicationFactors function and it is a data.table object with the deduplication factors for each region.
popRegFileName	The name of the file with the population counts for each region taken from a population register. It has 2 columns: region, N0.
pntRateFileName	The name of the file with the penetration rates for each region. It has 2 columns: region, pntRate.
regsFileName	The name of the .csv file defining the regions. It has two columns: tile, region. The first column contains the IDs of each tile in the grid while the second contains the number of a region. This file is defined by the user and it can be created with any text editor. It is required only for the state process negative binomial distribution.
gridFileName	The name of the .csv file with the grid parameters. It is required only for the state process negative binomial distribution.
rel_bias	The value of the relative bias for the population density of each region. The default value is 0.
cv	The coefficient of variation for the population density of each region. The default value is 0.

Value

A data.table object with the following columns region, omega1, omega2, pnrRate, regionArea_km2,
N0, dedupPntRate, alpha, beta. If regsFileName and gridFileName are not NULL the result will have 3 more columns:region, omega1, omega2, pnrRate, regionArea_km2, N0, dedupPntRate,
alpha, beta, theta, zeta, Q. They are needed only for the state process negative binomial distribution.

References

https://github.com/MobilePhoneESSnetBigData

bogdanoancea/inference documentation built on Nov. 28, 2020, 9:20 p.m.