syllabus‎ > ‎R: Intro‎ > ‎

Kagiso_Ndlovu 16121147


Visualization

The topic is elaborately covered. Clear definitions made and how the subject relates to background information. Meeting the ever changing user requirements continues to be the major problem, how can this be properly/adequately met in this field? Are there any techniques/tools adopted in the visualization field to cater for the color impaired individuals?

Polaris

Polaris application tool for visualization of relational multidimensional databases,leaves out non relational databases. For database tables to be effective in displaying multidimensional data, they should posses certain characteristics in them. Its appears it takes more than the data itself but how it is structured also contributes a lot to how its visualized by Polaris. That brings forth the question as to whether this would be time consuming for visualization process to complete compared to cases where only real data is used leaving out metadata? To graphically visualize multidimensional databases, analysts require some level of skill especially when dealing with Table Algebra.

The graphic file displayed by the Polaris tool can be determined by the cardinality of a record used. When the cardinality of the record is set to one, the graphics are simple bar charts or dot plots, and when the cardinality is greater than one, additional structures may be introduced to accommodate the additional records (eg. Stacked bar charts).

Domain: US Geographic Data - 1990 Census

Hypothesis:
States in the Central part of the American continent have a higher population and the higher the population, the bigger the area size of the state.

The above mentioned hypothesis is a prediction about the US geographic data gathered from a census in 1990. To test for this hypothesis, a graphical representation of all the states in the American continent was visualized in a map with all the different states color coded. A map key was displayed to show how the different states match to different color codes and population sizes. The size of the color coded circle correlates with the state population size. An interactive feature was added to the map such that once a given state is selected, then all of its geographical information including the state name, postal abbreviation, area size, population, are displayed on a table below the map.

The results from the visual representation of the census proved the hypothesis to be wrong. The states in the Central part of the continent are not densely populated as hypothesized compared to those in the Eastern part which are at the same time small in area size.

The data set could be further related to life expectancies from the different states in the US.

Code Wrapper










Comments