In a sense, data preparation is similar to washing freshly picked vegetables in so far as unwanted elements, such as dirt or imperfections, are removed.
Together with data collection and data understanding, data preparation is the most time-consuming phase of a data science project, typically taking eighty percent and even up to ninety percent of the overall project time. Automating some of the data collection and preparation processes in the database, can reduce this time to as little as 50 percent (This time savings translates into increased time for data scientists to focus on creating models).
To continue with our cooking metaphor, we know that the process of chopping onions toa finer state will allow for its flavours to spread through a sauce more easily than that would be the case if we were to drop the whole onion into the sauce pot.
Similarly, transforming data in the data preparation phase is the process of getting the data into a state where it may be easier to work with. Specifically, the data preparation stage of the data science methodology answers the question: What are the ways in which data is prepared?
To work effectively with the data, it must be prepared in a way that addresses missing or invalid values and removes duplicates, toward ensuring that everything is properly formatted.