Graphical representation of data through SPSS & Excel
Today was a day filled with graphics, trying to make it look interesting and easy to analyze. Initially we made some really sad graphs like the 2D bar graph, with data from the retail data set. Then we moved up slowly to 3D graphs and then the whole thing got really interesting when the 3D graph could be rotated. It is called the Interactive Graph where we take two attributes and then plot them on the graph.
Like sir explained in the class, Imagine that the OLAP cube is literally a cube and u cut a small portion of that cube for analysis. Here also we do the same wherein we segment the data into smaller cubes and analyze it from various angles. The process is called Online Analytic Process because we can quickly check up the details and get instant online results.
In order to make OLAP cubes, we should understand summary variable and grouping variable.
Summary variable is also called as scale. It is discrete i.e. round it off (if you are of the age 14.4 you tend to say 14). Here, age, monthly bill, SMS bill etc are considered. Grouping (Category) variable include gender, level of education, name of the service provider etc. You take both of these on the x axis and the y axis.
Although it was a disappointment to know that OLAP cube was not really a cube, it is a table which represents the data in a classified manner. Once the attributes are selected, a report will be generated which can be filtered according to the choice of comparison that the analyst requires. Along with this we also learned to analyze Bubble Graph, an example to this was the ICICI bubble graph analysis regarding Hedge funds, liquid funds, income etc. Then the report from a government website showed moving bubbles representing the no. of people employed and the trends in productivity and value added.
This type of graph is mainly used when we have more number of attributes. This is like Perception Mapping. In order to analyze and understand we should look into different attributes, as they are differentiated using different colors, from the closeness of it from the center to the outside. The further it is from the center the better it is. It helps in looking at the attributes from different perspective and interprets the data deeply
Author : Mohamed Sahle
Group: Marketing 5