Decomposition: We separated all separable features of our project (input new data --> dataInput, get stations and their km-numbers --> station_explorer, get line numbers --> line_finder, visualise the data on a map --> dashboard) into separate coding files to keep our code organized and reusable.
We changed the structure of our code to explore our data with the file station_explorer: By converting this code into a class we separated the tasks of "getting the km-number of a station on the line" from "getting the stations of a given line". So now, a function of this class does one precisely defined task (abstraction) and we can reuse the different parts if needed somewhere else in our code (decomposition).
Similarly to the change above, in the line_finder, we separated the task of "getting the lines of a given station" from the task of "finding common lines between two given stations".
Note that for better understanding, better overview and to be able to use only parts of this functions somewhere else, we decided to separate the feature of looking for information about stations for a given "line" (station_explorer) from the feature of looking for information about lines for a given "station" (line_finder). This enhances the aspect of decomposition.
We added much more comments and docstrings to our code than what we had before (to enhance abstraction). So now we can handle the pieces of our code more easily like little "black boxes", since we know what they contain, but don't have to understand the code each time we reuse it. (An example for this can be found in the station_exporer, the line_finder and the train_network.)