- Category: Uncategorised
- Written by Super User
- Hits: 61
Sometimes, I wonder, what to look for when doing data analytics. At least, in 2 data analytics classes which I attend,
I raised the same questions.How do we start with data analytic?
1. Building the objective, hyphothesis then drive into the data?(Top down approach) OR
2. Drive directly to the data, then find out what are the issues from multiple perspective to come out with analysis? (Bottom up approach)
For example, currently, I'm in having trouble to decide on what way to provide insight for preparing the analysis used for coming out with the audit business plan next year. We know that insight produced before the audit plan is quite difficult and tricky because we don't know for sure if there is any issues in the proposed audit plan. By just doing analyst and logic, sometimes we are not for sure the outcome. The whole objective of preparing the analysis is to at least direct and align the audit plan.
I've few method planned and proven to provide some insight to the analysis that I've done for the BP.
1. fivenum(x), which provide the mean,max,min,median,1st quartile and 3rd quartile. This is very useful and I used boxplot to populate the information.
2. another is to show category, then used histogram. Plus, if there is a different type, used facet_wrap. It's good.
Still there are still other method which I'm still exploring....