capm

Introduction

capm is a package for R, a free software environment for statistical computing and graphics.

capm stands for Companion Animal Population Management and is an initiative to guide and automate quantitative analysis to support companion animal population management.

Current version offers functions that allow users to implement the workflow outlined in Figure 1. The basic idea is simple, some functions are to implement a data collection process and other functions are to analyse collected data. Information generated in the analysis process support the design, implementation and assessment of population management programs. Detailed descriptions of workflow steps and used methods will be added to the links at the end of this page.

Figure 1. Workflow supported by the capm.

Data collection

Data collection is the first of the two main components of the workflow. When data is collected in censuses surveys, this component must be omitted. Remember that a census is a survey that includes ALL elements of the study population. For example, a household survey in a city is a census only if ALL households in that city are included in the survey. Costs associated with censuses are often prohibited and thus, sample surveys are the most common alternative to collect population data. The Data collection component is intended to support sample surveys.

The sample is a subset of population elements. Because we want information of the whole population, we must to infer that information (population parameters) from the sample. This process incorporates errors. We can not know the exact values of population parameters but we can get approximations (estimates). Estimates can be misleading but reasonable and useful estimates can be generated if a sound statistical process is followed to calculate the sample size and to select sampling units (elements included in the sample). To calculate the sample size, it is necessary to have a preliminar estimate of the parameters to be estimated. This preliminary estimates can be obtained in pilot studies or from previous studies. Using the capm, sample surveys for companion animal populations can be easily designed and implemented following well known statistical processes but without need to worry about them.

Data analysis

With sample or census data, some analysis can be made to support population management programs. For this second component, the capm has functions to estimate population parameters from sample data, according to the data collection process. Also, it is possible to build population pyramids conditioned not only on sex and age but on other characteristics such as reproductive status. The estimates of population parameters can be used as inputs in mathematical models to simulate population dynamics, prioritize interventions and build scenarios to assess the interaction between some interventions. The use of these models is straightforward in the capm.

Installation

To install the development version, run install_github("oswaldosantos/capm") in (ideally) RStudio. devtools package must be previously loaded.

To install the production version, run install.packages("capm") in (ideally) RStudio.

Gallery

In the data collection process, selected sampling units can be mapped within R or in external software such as Google Earth or QGIS, using files exported by the capm.

R

R

Google Earth

In the data analysis process, population pyramids and simulations of population dynamics help to characterize populations.

Population pyramid

Simulation of population dynamics (point estimates)

Global sensitivity of non sterilized stray dog population to all parameters

Global sensitivity of non sterilized stray dog population to each parameter

Local sensitivity of non sterilized stray dog population

Scenarios

Contributions

If you find any error or have suggestions to improve or include functionality to the capm I will glad to know it. Please get in touch so I can better understand your needs.

Documentation

Research documents

Defining priorities for dog population management through mathematical modeling: manuscript describing the model implemented in the function SolveIASA. It was drawn from the thesis' chapter 5.

Dog and cat management through sterilization: Implications for population dynamics and veterinary public policies: manuscript describing a practical application of the model described in the above manuscript.

Acknowledgments

I am particularly grateful to Fernando Ferreira and Marcos Amaku for their methodological advices.