Impressions

――I will write three things that I felt by explaining up to the 10th question of Data Science 100 Knock (Structured Data Processing).

Pandas is very useful for "conditionally extracting structured data"

--Structured data is ** a tabular summary of all the information you think you need ** ――You need to extract the information you want to see according to your purpose. ――For that purpose, I extract by specifying the condition, but ** Pandas is very convenient in that I can write the code of the condition specification concisely **

The point is "what kind of condition" and "what kind of code should be used"

――A feature of Pandas is that the condition specification and the code are paired. --For example, the condition specification "display the first 10 items" is expressed by the code ".head (10)". ――The shortcut to learning is to know this pattern, deepen your understanding through 100 knocks, and repeat until you can use it. -** Learning condition specifications and code pairs is the key to learning Pandas **

In practice, it is important to get into the habit of always thinking about "for what" and "what kind of data you want to extract".

――However, ** It cannot be used in practice just by being able to use it ** ――For example, in order to achieve the goal of "I want to increase sales by 5% year-on-year" as if I were a convenience store manager, "Place the products that are often bought as a set on the same shelf and set the purchase unit price. Suppose you have a strategy of "raise" ――In this case, since the purpose is clear, you can think that you should extract data such as "who", "when", "what product", and "how much" you purchased (* Of course). You can't just extract the data, you can't reach your goal without analyzing the data and getting strong suggestions.) ――However, if the purpose is unclear, no matter how much data can be extracted, it will not function as a means to solve business problems. -** Enjoy 100 knocks, always keeping in mind that "extracting data (+ analyzing data)" is one of the means to make good decisions. !! ** **