My 1st Dashboard :)



15 buildings, 35 aspects per building, each weighted according to importance, and then condensed into 6 key categories, with colour coding depending on how the results tumble out.

What helped me get the final result?

  • Pandas
    • to import data from spreadsheet
    • extract only the data I need
    • manipulate it so we end up with scores (0 – 1) instead of qualitative assessments like “in progress”
    • finally pivot the data into a summarized format
  • Matplotlib.GridSpec
    • to draw a grid into which I place the graphs
    • to easily achieve the correct spacing and placement between graphs
  • Matplotlib.PyPlot
    • a function to draw each graph and colour-code the bars instead of writing my code 15 times – don’t laugh, I’m really proud of this part: it was challenging for me!
    • and then to draw the actual graphs
  • Normal Python
    • for everything else like making sure the graphs appear in descending order of building size

What was difficult about it?

I ended up with different lists, dictionaries, and dataframes for different aspects of the project – the most challenging thing at the beginning was wrapping my head around extracting a value from one source, and passing it to another source as a selection criterion.

What was useful in overcoming the difficulties?

  • In trying to make my code as generic as possible I quickly discovered the merits of .iloc() over .loc() – especially in my for loops!
  • For a df with index names df1.index[0] is useful for getting the index names without referring to them by name!
  • Which would then allow me to get the column number based on the named index value in a related dataframe: df2.columns.get_loc(df1.index[0])
  • And this would allow me to extract values based on numeric locations: df2.columns.values[df1[0]]
  • And then zip was great for quickly building a dictionary from 2 lists: my_dict = dict(zip(my_list1, my_list2))

Comments are closed.

Create a website or blog at

Up ↑

%d bloggers like this: