Making sure data is clean and ready to use is a task that can't be overlooked. The following is an example of a very typical sales data set that had some format issues and some empty cells (null values).  While in some cases, we could process data as we receive it, any conclusions we might try to draw from it, would most likely be problematic  and misleading. In other words, useless.An initial exploration of the data allow us to see what do we need to fix. In this case, I first inspected the data using common commands such as data.head (10) to mainly see the column headers and whether there were integers, floats or strings. When dealing with big tables it is recommended to see the first rows and the last ones instead of loading all the rows as this might slow down the tools being used. At this point we can also see if the are other problems like empty spaces or some duplicates. We can also see how many columns and rows are there.

After sorting, filtering, renaming, and fixing the quirks in the dataset, the data was finally ready to reveal its secrets. I performed some brief calculations to see the total sales per region, the average price per product, and how many orders each customer had placed. Surely, this was enough to start seeing exactly how well the business is doing!

GitHub Repository  Original files are available here 

Google Colab here