Autoformatting python source code

[kdheepak]

Hi! Firstly, thanks for developing and maintaining this project!

I wanted to check if you'd be interested in using an autoformatter for your source code?

At the moment, there's a number of lines that exceed conventional and unconventional character limits in Python projects (80 characters, 120 characters or even 150 characters per line) e.g.:

openTEPES/openTEPES/openTEPES_InputData.py

Lines 566 to 568 in 9dd774b

	mTEPES.n2 = Set(initialize=mTEPES.nn, ordered=True , doc='load levels' , filter=lambda mTEPES,nn : nn in mTEPES.nn and sum(pDuration [p,sc,nn] for p,sc in mTEPES.ps) > 0)
	mTEPES.g = Set(initialize=mTEPES.gg, ordered=False, doc='generating units' , filter=lambda mTEPES,gg : gg in mTEPES.gg and (pRatedMaxPowerElec [gg] > 0.0 or pRatedMaxCharge[gg] > 0.0 or pRatedMaxPowerHeat [gg] > 0.0) and pElecGenPeriodIni[gg] <= mTEPES.p.last() and pElecGenPeriodFin[gg] >= mTEPES.p.first() and pGenToNode.reset_index().set_index(['index']).isin(mTEPES.nd)['Node'][gg]) # excludes generators with empty node
	mTEPES.t = Set(initialize=mTEPES.g , ordered=False, doc='thermal units' , filter=lambda mTEPES,g : g in mTEPES.g and pRatedLinearOperCost[g ] > 0.0)

This requires scrolling left to right on GitHub and in other text editors (where soft wrapping lines are not enabled) and makes it harder (in my humble opinion) to read some parts of the code.

There's also a number of lines where manual formatting of the spaces is used for aligning the code:

openTEPES/openTEPES/openTEPES_InputData.py

Lines 407 to 429 in 9dd774b

	#%% generation parameters
	pGenToNode = dfGeneration ['Node' ] # generator location in node
	pGenToTechnology = dfGeneration ['Technology' ] # generator association to technology
	pGenToExclusiveGen = dfGeneration ['MutuallyExclusive' ] # mutually exclusive generator
	pIndBinUnitInvest = dfGeneration ['BinaryInvestment' ] # binary unit investment decision [Yes]
	pIndBinUnitRetire = dfGeneration ['BinaryRetirement' ] # binary unit retirement decision [Yes]
	pIndBinUnitCommit = dfGeneration ['BinaryCommitment' ] # binary unit commitment decision [Yes]
	pIndBinStorInvest = dfGeneration ['StorageInvestment' ] # storage linked to generation investment [Yes]
	pIndOperReserve = dfGeneration ['NoOperatingReserve' ] # no contribution to operating reserve [Yes]
	pMustRun = dfGeneration ['MustRun' ] # must-run unit [Yes]
	pInertia = dfGeneration ['Inertia' ] # inertia constant [s]
	pElecGenPeriodIni = dfGeneration ['InitialPeriod' ] # initial period [year]
	pElecGenPeriodFin = dfGeneration ['FinalPeriod' ] # final period [year]
	pAvailability = dfGeneration ['Availability' ] # unit availability for adequacy [p.u.]
	pEFOR = dfGeneration ['EFOR' ] # EFOR [p.u.]
	pRatedMinPowerElec = dfGeneration ['MinimumPower' ] * 1e-3 * (1.0-dfGeneration['EFOR']) # rated minimum electric power [GW]
	pRatedMaxPowerElec = dfGeneration ['MaximumPower' ] * 1e-3 * (1.0-dfGeneration['EFOR']) # rated maximum electric power [GW]
	pRatedMinPowerHeat = dfGeneration ['MinimumPowerHeat' ] * 1e-3 * (1.0-dfGeneration['EFOR']) # rated minimum heat power [GW]
	pRatedMaxPowerHeat = dfGeneration ['MaximumPowerHeat' ] * 1e-3 * (1.0-dfGeneration['EFOR']) # rated maximum heat power [GW]
	pRatedLinearFuelCost = dfGeneration ['LinearTerm' ] * 1e-3 * dfGeneration['FuelCost'] # fuel term variable cost [MEUR/GWh]
	pRatedConstantVarCost = dfGeneration ['ConstantTerm' ] * 1e-6 * dfGeneration['FuelCost'] # constant term variable cost [MEUR/h]
	pLinearOMCost = dfGeneration ['OMVariableCost' ] * 1e-3 # O&M term variable cost [MEUR/GWh]
	pOperReserveCost = dfGeneration ['OperReserveCost' ] * 1e-3 # operating reserve cost [MEUR/GW]

I really like how much consistency you've enforced in aligning the whitespace like this, it makes it easy to visually parse the structure of various sub-expressions in the code.

However, manually aligning the whitespace can be time-consuming. Changing a parameter name, or adding or removing an existing one will require a modeller to manually align these parameters again in order to continue maintaining consistency. This will increase the diff every time a modeller makes a contribution.

Furthermore, if a modeller were to make changes that they wish to contribute back to this project, the manual alignment of code like this can increase the chances of conflicts that would need to be resolved.

Would you consider using an autoformatter (black, ruff, yapf, autopep etc are commonly used ones) such that it autoformats the code when you save or when you commit code to a project? A number of these formatters have various options you can configure too, that you can tweak to your liking.

In my humble opinion, it would make it more easier to anyone to upstream changes or contributions.

But no worries if you choose not to adopt autoformatting! I know that the current style can make it very easy to read the code and there are certainly clear benefits to the current approach that you'd lose if you were to choose to adopt an autoformatter.

If you think this is not a direction you'd like to take the project, feel free to close the issue :)

[arght]

Thanks for your comments. We will check the potential use of autoformatter for improving the code aspect while facilitating the contributions.