Downloads |
Epidemiology Datasets
The Garki Project
A pdf version of the book The Garki Project (Molineaux L, Gramiccia G, 1980. The Garki Project. Geneva: World Health Organisation) is available for download from http://whqlibdoc.who.int/publications. The links below provide access to the dataset and to scanned versions of the maps of the study villages.
Garki Datasetbr
* Access DB (4.8 MB)
* Foxpro (*.dbf) Tables (4.6 MB)
* SQL (mysql dump) (4.2 MB)
Scanned Maps of the Garki Villages
* Village Cluster 1 (10.3 MB)
* Village Cluster 2 (15.6 MB)
* Village Cluster 3 (16.8 MB)
* Village Cluster 4 (23.7 MB)
* Village Cluster 5 (19.7 MB)
* Village Cluster 6 (9.5 MB)
* Village Cluster 7 (14.6 MB)
* Village Cluster 8 (12.3 MB)
Documents
Defining clinical malaria in vaccine trials
Software
As part of our work on statistical methodology the Biostatistics Unit writes software for specialist statistical applications, mostly in parasite epidemiology.
Several of these applications are written in the free software Winbugs, obtainable from the MRC Biostatistics Unit, Cambridge, England. The code in the .odc format used by Winbugs, can be downloaded here. General information about how to use Winbugs is available from the Winbugs website. A tutorial for new users can also be downloaded (http://www.mrc-bsu.cam.ac.uk/bugs/winbugs/winbugs-demo.pdf).
Users should contact Tom Smith by email if they have specific questions about how to use the software which we have written.
Maximum likelihood estimation of allele frequencies for malaria parasites
Download Fortran source code, and compiled version for Windows (zip-file)
Latent class model for non-parametric resolution of a two component mixture, with one component sampled directly.
- Estimation of attributable fractions in malaria and modelling the relationship of attributable morbidity to parasite density.
- Determination of the percentage of samples positive in immunoassays
Vounatsou P, Smith T, Smith AFM. Bayesian analysis of two component mixture distributions: application to the estimation of malaria attributable fractions. Appl Stat. 47 (4), 575-587, 1998.
Download Winbugs code (zip-file)
Download documentation (pdf-format)
Two-state Hidden Markov Model with covariates
Estimation of force of infection and duration of infection in malaria from panel data of parasite typing
Smith T, Vounatsou P. Estimation of infection and recovery rates for highly polymorphic parasites when detectability is imperfect, using hidden Markov models. (submitted for publication)
Download Winbugs code (zip-file)
Download documentation (pdf-format)
Regression model with Poisson errors in both x and y.
Calibration of different trapping methods for mosquitoes Hii JLK, Smith T, Mai A, Ibam E, Alpers MP. Comparison between anopheline mosquitoes (Diptera: Culicidae) caught using different methods in a malaria endemic area of Papua New Guinea. Bull Ent Res. 90: 211-219, 2000
Download Winbugs code (zip-file)
Download documentation (pdf-format)