Monday, October 6, 2014

Let us understand different charts in an interesting way. I have chosen my favorite film star's film career to understand the charts and Analytics with qlikview

Prakash Raj

Prakash Raj is an Indian Actor, who is a favorite of many. He is versatile to an extent that he can act as someone who loves an actress in one film and as a loving dad of the same actress in the next film; whats to be noted is that he acts very convincingly in both the roles to make both the characters become a brand in itself.

I have made the below analysis of his filmography with the data available in wikipedia. I am sorry if any information available in the link is wrong.  This is just the visual representation of available data to understand the usage of graphs and qlikview.

Data got from:

The above is a line chart. This kind of chart can be used when the growth of a measure needs to be studied from the chart. From the above chart we can say that Praash sir has acted maximum number of movies in 2011 and 2006. We can also say that he was on the run, too busy, working as hard as possible day night between 2004 and 2006 (But the contrary doesnt hold good. If the number of films is less we cant say he didnt work day night as it depends on the movie itself. The story of the movie is a variable factor which cant be incorporated to our analysis) .

This lines chart has just one dimension and one measure:
Dimension: Year
Measure: dual(Language&'('&count(Film)&')',count(Film))

Now, lets discuss each of the above charts.

Above is a pie chart. This can be best used when we have just one measure and one dimension (Though qlikview allows us to use for more dimensions, i would recommend other charts for more dimensions). A Pie chart can be used when an overall picture of the measure is to be seen for a dimension at a glance.

Inference: Sir has acted 296 films in his career (acc to the link we use for data), which belong 5 languages. The number of films in each language can be seen in the pie chart.

2 dimensions used in line chart: The above line chart has two dimensions. Here we can see the trend in number of films for various languages across years.
Dimensions: Language and Year
Measure: Count(Film)

Year is given in X axes while the other dimension, language, is differentiated by different colous according to the legend.

This is again a line chart with two dimension. But the difference is, the expression uses set analysis.

Problem statement for the chart: Awards of Prakash sir, with years and languages as dimension.

We can infer from the above graph that prakash sir has won max number of awards in 2003 with 4 for Telugu movies.

We dont have a separate table for awards in the page I have used. So, how to get this details? We have a data field called Notes which has mentions about awards and a lot of other things, like if it is a directorial venture. So we use in the expression a set analysis.


If we give Count(Film) this chart will be exactly like the previous one with the chart drawn considering all his films.
{} is the set analysis used.
Here we count(Film) only for those records which has '*Awards*' as notes.
*Awards* - used to find if notes has Awards preceded and followed by any characters. Substring.

The syntax for set analysis is

The above is a bar chart.  The first one depicts the stats of his number of movies according to first letter of the film name and the second depicts the award winning movies with the first letter of film name.

In this chart we are using expression as dimension.
Dimension: Left(Film,1) this will take the first letter of the films as dimension
Expression: Count(Film)

This doesn't help much in analysing sir's filmography but might be helpful in other scenarios. Let's take this example just to learn expression as dimension and sorting.

To sort your results based on the expression value, just copy paste the entire expression in the sort expression of the dimension.

Count(Film) is my expression. I have just pasted it in the sort of my dimension to get a chart sorted like above.

I hope this article was helpful in understanding Line chart with mutiple dimensions and with set analysis. Have fun working on such a dashboard for your favorite film star; learn having fun.

I hope the details in wiki about Prakash sir is right and ,hence, hope my analysis is right.

0 Scribbles:

Post a Comment

Thank You for taking pains, commenting :)


Search This Blog