% Rapport package team % Multidimensional Scaling % 2011-04-26 20:25 CET ## Description In this template Rapporter will present you Multidimensional Scaling. ### Introduction [Multidimensional-scaling](http://en.wikipedia.org/wiki/Multidimensional_scaling) is a technique which gives us a visual representation about the distances between the observations. ### MDS Below you can see a plot, that presents you the distance between the observations, which was calculated based on _Age_, _Internet usage for educational purposes (hours per day)_ and _Internet usage in leisure time (hours per day)_. [![](plots/MDS.tpl-1.png)](plots/MDS.tpl-1-hires.png) ##### What can be seen here? ###### Outsiders 84 differs the most from the others, and 8 seems to be the most "common" observation, which lie nearest to all other observations. ###### Outsider Pairs _284_ and _84_ (8.02) are the "furthest", _280_ and _1_ (0) are the "nearest" to each other. ###### In General Now let's see which observations can be said statistically far/similar to each other in general. The _16_ pairs with the biggest differences and the _10_ pairs with the smallest differences will be presented. In the brackets you can see the amount of the distances between two observations. There are _17_ observations which are the most similar, and equal in the same time, that is a higher number than the wanted _16_, thus will not be reported one-by-one. Set _17_ as parameter _max.dist.num_ to check the pairs if you are interested. There are _318_ observations which are the most similar and equal in the same time, that is a higher number than the wanted _10_, thus will not be reported one-by-one. Set _318_ as parameter _min.dist.num_ to check the pairs if you are interested. ## Description In this template Rapporter will present you Multidimensional Scaling. ### Introduction [Multidimensional-scaling](http://en.wikipedia.org/wiki/Multidimensional_scaling) is a technique which gives us a visual representation about the distances between the observations. ### MDS Below you can see a plot, that presents you the distance between the observations, which was calculated based on _Age_, _Internet usage for educational purposes (hours per day)_ and _Internet usage in leisure time (hours per day)_. [![](plots/MDS.tpl-1.png)](plots/MDS.tpl-1-hires.png) ##### What can be seen here? ###### Outsiders 84 differs the most from the others, and 8 seems to be the most "common" observation, which lie nearest to all other observations. ###### Outsider Pairs _284_ and _84_ (8.02) are the "furthest", _280_ and _1_ (0) are the "nearest" to each other. ###### In General Now let's see which observations can be said statistically far/similar to each other in general. The _17_ pairs with the biggest differences and the _30_ pairs with the smallest differences will be presented. In the brackets you can see the amount of the distances between two observations. According to the used variables (_Age_, _Internet usage for educational purposes (hours per day)_ and _Internet usage in leisure time (hours per day)_) the _17_ furthest pair of observations are: * _284_ and _84_ (8.02) * _224_ and _84_ (7.87) * _230_ and _84_ (7.84) * _84_ and _68_ (7.81) * _463_ and _84_ (7.79) * _583_ and _84_ (7.79) * _582_ and _84_ (7.72) * _122_ and _84_ (7.72) * _460_ and _84_ (7.72) * _606_ and _84_ (7.7) * _607_ and _84_ (7.7) * _128_ and _84_ (7.69) * _253_ and _84_ (7.69) * _84_ and _41_ (7.69) * _269_ and _84_ (7.65) * _376_ and _84_ (7.63) * _506_ and _84_ (7.63) There are _318_ observations which are the most similar and equal in the same time, that is a higher number than the wanted _30_, thus will not be reported one-by-one. Set _318_ as parameter _min.dist.num_ to check the pairs if you are interested. ## Description In this template Rapporter will present you Multidimensional Scaling. ### Introduction [Multidimensional-scaling](http://en.wikipedia.org/wiki/Multidimensional_scaling) is a technique which gives us a visual representation about the distances between the observations. ### MDS Below you can see a plot, that presents you the distance between the observations, which was calculated based on _drat_, _cyl_ and _mpg_. [![](plots/MDS.tpl-2.png)](plots/MDS.tpl-2-hires.png) ##### What can be seen here? ###### Outsiders Honda Civic differs the most from the others, and Ferrari Dino seems to be the most "common" observation, which lie nearest to all other observations. ###### Outsider Pairs _Honda Civic_ and _Cadillac Fleetwood_ (5.48) are the "furthest", _Mazda RX4 Wag_ and _Mazda RX4_ (0) are the "nearest" to each other. ###### In General Now let's see which observations can be said statistically far/similar to each other in general. The _17_ pairs with the biggest differences and the _30_ pairs with the smallest differences will be presented. In the brackets you can see the amount of the distances between two observations. According to the used variables (_drat_, _cyl_ and _mpg_) the _17_ furthest pair of observations are: * _Honda Civic_ and _Cadillac Fleetwood_ (5.48) * _Honda Civic_ and _Lincoln Continental_ (5.39) * _Dodge Challenger_ and _Honda Civic_ (5.25) * _Toyota Corolla_ and _Cadillac Fleetwood_ (5.1) * _Toyota Corolla_ and _Lincoln Continental_ (5.04) * _Honda Civic_ and _Merc 450SLC_ (4.85) * _Fiat 128_ and _Cadillac Fleetwood_ (4.79) * _Honda Civic_ and _Merc 450SE_ (4.74) * _Honda Civic_ and _Duster 360_ (4.74) * _AMC Javelin_ and _Honda Civic_ (4.74) * _Fiat 128_ and _Lincoln Continental_ (4.74) * _Honda Civic_ and _Chrysler Imperial_ (4.68) * _Honda Civic_ and _Valiant_ (4.68) * _Honda Civic_ and _Merc 450SL_ (4.67) * _Dodge Challenger_ and _Toyota Corolla_ (4.67) * _Pontiac Firebird_ and _Honda Civic_ (4.52) * _Honda Civic_ and _Hornet Sportabout_ (4.46) According to the used variables (_drat_, _cyl_ and _mpg_) the _30_ nearest pair of observations are: * _Mazda RX4 Wag_ and _Mazda RX4_ (0) * _Chrysler Imperial_ and _Duster 360_ (0.08) * _Merc 230_ and _Datsun 710_ (0.13) * _Lincoln Continental_ and _Cadillac Fleetwood_ (0.13) * _Merc 450SL_ and _Merc 450SE_ (0.15) * _AMC Javelin_ and _Merc 450SLC_ (0.15) * _Pontiac Firebird_ and _Hornet Sportabout_ (0.15) * _AMC Javelin_ and _Chrysler Imperial_ (0.17) * _AMC Javelin_ and _Duster 360_ (0.19) * _Merc 450SLC_ and _Merc 450SE_ (0.2) * _Merc 280C_ and _Merc 280_ (0.23) * _AMC Javelin_ and _Merc 450SE_ (0.25) * _Merc 450SL_ and _Hornet Sportabout_ (0.28) * _Merc 280_ and _Mazda RX4_ (0.3) * _Merc 280_ and _Mazda RX4 Wag_ (0.3) * _Merc 450SLC_ and _Duster 360_ (0.3) * _Chrysler Imperial_ and _Merc 450SLC_ (0.31) * _Pontiac Firebird_ and _Merc 450SL_ (0.32) * _Merc 450SLC_ and _Merc 450SL_ (0.35) * _Toyota Corona_ and _Datsun 710_ (0.35) * _Toyota Corolla_ and _Fiat 128_ (0.36) * _AMC Javelin_ and _Merc 450SL_ (0.38) * _Merc 240D_ and _Datsun 710_ (0.4) * _Merc 450SE_ and _Hornet Sportabout_ (0.41) * _Chrysler Imperial_ and _Merc 450SE_ (0.41) * _Volvo 142E_ and _Merc 230_ (0.42) * _Merc 450SE_ and _Duster 360_ (0.44) * _Maserati Bora_ and _Camaro Z28_ (0.45) * _Toyota Corona_ and _Merc 230_ (0.46) * _Pontiac Firebird_ and _Merc 450SE_ (0.46) ------- This report was generated with [R](http://www.r-project.org/) (3.0.1) and [rapport](https://rapporter.github.io/rapport/) (0.51) in _3.338_ sec on x86_64-unknown-linux-gnu platform. ![](images/logo.png)