Cell 1:

• Do you need to specify the sign convention? I think it might be confusing as the users anyways input the costs with a positive sign.
• Maybe already define the meaning of the values Y, M, Q for the freq parameter here?

Cell 2:

• comment not matching code (20000 vs 50000 and 12000 vs 8000)

Cell 3:

• "Three methods are available to calculate cash flows:
• I would maybe repeat the parameters instead of the ... for clarity

Cell 5:

• Consider an option to add a row with totals in the dataframe format

Cell 6:

• ... draws a bar chart of the cash flows

Cell 7:

• Is it voluntary that you do not compare the same init_cost, periodic_cost and periodic_income between the with- and without-growth instances? See the comment on cell 2.

Cell 10:

• Not exactly sure of what is meant with "annualised" but it is probably because I am not familiar with the field :) I think I get the idea though

Cell 11:

• Do we care about the days when we consider the monthly case? What happens if we put 2022-01-15 instead of 2022-01-01 for the date? Maybe you could shortly comment on that (if relevant)

Cell 13:

• Maybe add an example of the manual merge of the custom flows?

Cell 15:

• The From YAML part seems to be incomplete here, but you are probably aware of it :)

Hi @luseverin,
Thanks for the review!

Here are some of the changes I made (also based on @chahank remarks):

• Improved the Quickstart to show more of the capabilities right from the start, with minimal amount of scrolling.
• Added meaning of periods aliases as well as url to pandas documentation
• Aligned the values with comment as well as for the different examples
• Improved description of the plotting function
• Improved details on sub-yearly period frequencies
• Improved description of the merging (also now custom cashflows are merged)

I think I adressed all other comments.
For the YAML part, I don't want the notebook to have to write a file, so the code is within the markdown, not as a python cell. I added a bit more to it.

I think also str or int or datetime object word, right?

from code below this should be int

you could add that all pandas period aliases are allowed?

see comment above, maybe adapt docstring if only ints work here

see type above

Just a note for compatibility with config, that you probably already are aware. Here is seems the config.custom_cash_flows must have cost and income columns, but in the config is seems only to have values which can be positive or negative:

This currently breaks (copied definition from other tutorial)

custom_schedule = [
    {"date": "2024-01-01", "value": -1000000},  # Initial cost
    {"date": "2029-01-01", "value": -200000},  # Mid-term overhaul
    {"date": "2034-01-01", "value": 500000},  # Terminal value
]

custom_finance = CostIncomeConfig(custom_cash_flows=custom_schedule)
ci = CostIncome.from_config(custom_finance)

is it on purpose that these parameters do not have type suggestions? Because there would be too many that work?

the single-date cash flows are here always computed with the start time of the period. Is this consistent with how they are computed? Could it not happen that sometimes a cash flow is left out if it is "in the middle" of a period?

the code seems to return tuple of axis (ax_bar, ax_net), which is not the same as plt.Figure?

It looks to me that this private method is not used anywhere? Will it be used somewhere else?

not clear if this is oblic or private function? looks private here but public in test?