Data Science has different approaches depending on what you are trying to do.
EDA(Exploratory Data Analysis) has a classification of how to view tabular data (table data). There are also classifications according to purpose such as "I want to make predictions" and "I want to prove from data". These are different approaches.
This page is classified according to the phenomenon to be treated. The difference between the three phenomena of "causality, system, and time" is the difference in the method of analysis.
As types of phenomena, I divided them into causality, system, and time. There may be others, but for now, I'll stick with these three.
"Causality" is also called "Cause and effect".
The word "Causality" sounds difficult, but it has become a daily topic of discussion, like "I think the reason is this" or "This is because it is XX".
When we look at things as a system, we think of the thing that is happening as the ``whole'', divide it into ``parts'', and think about the relationships between the parts.
Depending on how you think about "parts" and how you think about relationships, you can proceed in various ways.
Things happen in time.
When looking at the data, it becomes easier to express the relationships if you consider the order of time.
As a way of looking at phenomena, I divided them into three categories: causality, system, and time.
In the case of a narrow or uncomplicated phenomenon, focusing on just one of these may lead to a solution. On the other hand, when it comes to phenomena that occur in society and factories, it seems to be good to use and mix the three perspectives.
Different phenomena require different methods of analysis.
Data, mathematics, logic, and graphs are different for each analysis. There are relationships between these elements. For example, a scatter chart is a graph that visually examines the mathematics of correlation.
The figure below is just an example. I write what comes to my mind.
Mathematics and logic are treated as one in mathematics, but they are separated here. Mathematics is something that can be expressed in formulas, and logic is expressed in words, not formulas. In mathematics, formulas are like symbolic relationships, but in data science, the size of specific numbers in formulas and the meaning of categories are important, so this way of interpretation is used.