Description

Spatial data has become ubiquitous in a myriad of different disciplines and poses substantial challenges to both applied scientist and statisticians. Such data arises i.e. in  climatology or enviromental sciences (where different weather characteristics such as temperature, humidity etc are recorded at fixed locations), in economics or political sciences (where housing prices or election results are collected at ZIP level) as well as in  criminology and forestry (where the locations of crime events or tree stands are of interest).

Recently, spatial data on structured domains i.e. roads or railway systems or on the sphere have stimulated a immense interest. All of these examples form a particular type of spatial data, all of which will be covered in detail in class. The 2 plus 2 class is designed to provide a detailed treatment of all fields of spatial statistics, namely:

1) Geostatistics,

2) Spatial areal data, and

3) Spatial point patterns.

Starting with univariate techniques with focus on one outcome in space, techniques for multivariate geostatistical and areal data and also planar marked point patterns as well a point patterns on networks (e.g. car accidents) are discussed in detail. Exercises will cover both, calculations and practical applications of spatial statistics using the capacities of R. Further, a short introduction to geoinformation system (GIS) using the open-source solution QGIS is presented. Although a solid knowlegde of R is not necessary for this course, some prior knowledge in R basics is recommend to follow the exercises and applications. Access key will be provided in the first lecture. The registration to the moodle course is obligatory.