# Joint Dimension Reduction and Clustering in R

I am trying to understand the Joint Dimension Reduction and Clustering in R which comes in its new package clustrd where they have followed (this Article). The macro data from the package and its output are given below:

            GDP   LI   UR    IR   TB  NNS
Australia   4.8  8.4  8.1  5.32  0.7  4.7
Canada      3.2  2.5  8.4  5.02  1.6  5.2
Finland     3.9 -1.0 11.8  3.60  8.8  7.7
France      2.3  0.7 11.7  3.69  3.9  7.3
Spain       3.6  2.5 19.0  4.83  1.2  9.6
Sweden      4.1  1.1  8.9  4.20  7.0  4.0
USA         4.1  1.4  4.5  5.59 -1.4  7.0
Netherlands 2.9  1.6  4.2  3.69  7.0 15.8
Greece      3.2  0.6 10.3 11.70 -8.3  8.0
Mexico      2.3  5.6  3.2 20.99  0.0 12.7
Portugal    2.8 -7.5  4.9  4.84 -8.7 14.0
Austria     1.1  0.6  4.7  3.84 -0.6  9.4
Belgium     1.4 -0.1  9.6  3.64  4.5 12.4
Denmark     1.0  1.5  5.3  4.08  3.3  5.0
Germany     0.8 -2.0  9.5  3.74  1.5  7.7
Italy       0.9 -0.4 12.3  6.08  4.3  8.2
Japan       0.1  5.4  4.2  0.74  1.2 15.1
Norway      1.4  0.9  3.3  4.47  7.1 15.1
Switzerland 1.1  2.1  3.8  1.84  4.4 13.2
UK          1.2  4.9  6.4  7.70 -0.5  4.8


I'm very much clear about biplot but here I'm confused with the given plots. Would appreciate if someone give me point to interpret these plots. Thanks in advance for any help.

• I'm very much clear about biplot but here I'm confused with the given plots Sounds strange. Isn't the 1st one a biplot (with possibly some centering of the data and some form of scaling - the authors should have informed which)? Some form of clustering the countries points was done (which method? - consult with the source) and and the 2nd pic shows the 3 clusters' centroids profiles. Those 3 centroids poins were added to the biplot as - I expect - supplementary points. This is how it all looks to me at an instant glance. – ttnphns Nov 6 '16 at 14:01
• Are you asking more about what the plots might be, or asking about the analysis algorithm (described in the linked article) which has lead to the plots in the end? – ttnphns Nov 11 '16 at 14:17
• I am looking for interpretation of these graphs. – MYaseen208 Nov 11 '16 at 14:20
• Then see my comment above. The first pic is a biplot of the row and column items of the data in a space of (some) principal latent axes. Clusters of row items are also added there, as points (C1,C2...) The second pic looks like profiles of the clusters in the space of the data column items (i.e. characteristics). I presumed you know how to interpret a biplot or a line plot. – ttnphns Nov 11 '16 at 14:31