The ability to link notebook cells and switch between SQL and Python gives analysts the freedom to decide when to use which tool. "The best tool for the respective subtask" should be the motto. As with business intelligence tools, the use of standard components is always preferable, for example in data visualization. So if a readable bar chart can be created easily and satisfactorily using the on-board tools of a notebook, it is not necessarily necessary to write Python code for this purpose. If the analysis requirements go beyond this or if the use of special algorithms is planned, reaching into the extended toolbox - called Python - is probably the solution. As a continuation of the analysis path started in Fig. 3, Fig. 4 illustrates the further progress. As algeria telegram screening part of the exploratory data analysis, for example, the data visualization library Seaborn was used in a Python cell to create a boxplot diagram [9] . With the standard diagram types provided by the manufacturer, this representation would probably not have been possible at all or only with tricks.
If you enter the world of machine learning as a business analyst, you will quickly see the power of Python as a programming language and the associated ecosystem with its large selection of software libraries. Fig. 4 shows an example of how time series are visualized using the Pandas , NumPy and matplotlib libraries mentioned at the beginning and how a dataset is prepared for subsequent ML model training using the Prophet library . This type of task can be implemented very effectively with just a few lines of code and clearly shows the complementary application area of Python compared to SQL.