[R-sig-Geo] How to fit Spatial logistic regression models to DHS data

[Bedilu Ejigu]

I am analyzing geospatial data come from malaria intervention survey,
to compare standard multilevel models with spatial models.   Some of
the variables in my dataset are the following:



1.      malaria-malaria test result(1-presence, 0-absence) which is
our outcome variable

2.      LATNUM-coordinates of the survey cluster

3.      LONGNUM- coordinates of the survey cluster

4.      hv024-region (categorical variable)

5.      hv025-residence (urban/rural)

6.      hv227 -net use (yes/no)

7.      hv270 -wealth index(poorest, poorer, middle, richer, richest)

8.      hc1 – age in days

9.      hc27- sex (male/female)

10.    hc68-educational level (no education, primary, secondary)

11.    anebin- Anemia level(1-anemic,0-nonanemic)





 What I want to fit is a spatial logistic regression model by using
the aforementioned variables using any of the packages in R which can
handle the task (i.e. prevMap, geoRglm).  Can anyone help me on how to
fit such a spatial logistic regression model? If possible, and someone
did similar tasks before, could you share me your R code?



 Sample dataset, which shows the structure of my dataset:



hv024

hv025

hv227

hv270

hc1

hc27

hc68

LATNUM

LONGNUM

anebin

malaria

western

rural

yes

middle

18

female

middle/jss/jhs

5.076585

-2.88716

0

0

western

rural

yes

poorer

42

female

middle/jss/jhs

5.076585

-2.88716

0

0

western

rural

yes

poorer

15

male

middle/jss/jhs

5.076585

-2.88716

1

0

western

rural

yes

poorer

30

male

middle/jss/jhs

5.076585

-2.88716

1

0

western

rural

yes

middle

39

male

primary

5.076585

-2.88716

0

0

western

rural

yes

middle

19

male

primary

5.076585

-2.88716

1

0

western

rural

no

poorer

28

male

no education

5.076585

-2.88716

1

0

western

rural

no

poorer

8

male

primary

5.076585

-2.88716

1

0

western

rural

yes

middle

32

male

no education

5.076585

-2.88716

1

0

western

rural

yes

middle

59

male

middle/jss/jhs

5.076585

-2.88716

0

0

western

rural

yes

middle

40

male

NA

5.076585

-2.88716

1

0

western

rural

yes

poorer

36

male

middle/jss/jhs

5.076585

-2.88716

0

0

western

rural

yes

poorer

19

male

no education

5.076585

-2.88716

1

0

western

rural

yes

poorer

19

female

NA

5.076585

-2.88716

1

0

western

urban

yes

richer

9

female

middle/jss/jhs

5.286215

-2.76342

0

0

western

urban

no

richest

48

female

primary

5.286215

-2.76342

0

0





With best regards,



 Bedilu


*_______________________________________________*

	[[alternative HTML version deleted]]