Table 1.
Ten numerical variables included in the analysis.
Table 2.
Four categorical variables included in the analysis.
Fig 1.
Factor analysis for mixed data results.
A) The strength (length of arrow) and direction (direction of arrow) of correlation of each numerical variable with respect to the first and second dimensions. B) All variables, both numerical and categorical, and their relationship strength to the first two dimensions. C) The first eight dimensions extracted using FAMD and their respective explanation of data variability.
Fig 2.
Clusters resulting from hierarchical clustering analysis.
Each point in the map represents one country-year observation (one country in one year), which cluster it belongs to, and where it falls with respect to the first two dimensions. The data points are color-coded according to the clusters to which they belong. The size of a data points reflects the level of outgoing refugee flows from the corresponding country in the corresponding year. Note that the refugee data is used here for visualization purposes only and not part of the FAMD analysis.
Fig 3.
Conflict category mapped against the first and second dimensions.
Each point represents an individual country-year observation, color-coded according to its level of conflict. Labeled category points correspond to the center of gravity of observations belonging to that category.
Fig 4.
Seven countries and their movement through the first two dimensions in the years of the dataset.
For points that are closely clumped together, their centers of gravity were used to ease readability. The interested reader is referred to S1 Table for the cluster memberships of all countries from 2000 to 2015.