hemo_db.R
-> import and cleaninghemo_bygroup.R
-> cross-sectional analysis using baseline datahemo_byvisit.R
-> longitudinal analysis using folow-ups!
- definir variable dependiente
- Var. dependiente = cada una de las mediciones hematológicas
- Y: gaussian distribution identity link function
- definir variables independientes
- Var. independiente [time-invariant] = infeccion (especie de plasmodium, dicotomica)
- Var. independiente [confusores] = edad (continua), sexo (dicotómica)
- objetivo:
- test if, in average, the change of Y in time is related with a time-invariant exposure
- this translates as including an interaction term between Y and the time-invariant exposure (Vittinghoff, 2nd edition, p271)
- correlation structure
- the repeated measurements were taken through time
- only three time measurements per pacient were included
- the measurement interval times are not equally spaced
- autoregressive process 1 (ar1) because the correlation between times differ
- clustered data
- each pacient had 3 visits each with one measurement for all the haematological variables
- dataset is in a long (tidy) format
20180103
-
recepción de las bases
-
estructura:
- Controles (n=308) del año 2010
- Pf positivos (n=34) años 2011 y 2012
- Pv positivos (n=93) año 2011
20181201
- importar y limpiar base de datos
20190506
- union de base de datos y ejecución de análisis descriptivos
20200204
- corrección por autocorrelación debido a mediciones repetidas
@software{andree_valle_campos_2020_4014205,
author = {Andree Valle Campos},
title = {avallecam/hemogr: First release},
month = sep,
year = 2020,
publisher = {Zenodo},
version = {v0.1},
doi = {10.5281/zenodo.4014205},
url = {https://doi.org/10.5281/zenodo.4014205}
}