my n°1 programming tip

i recently read an article titled “my top 1 programming tip ever”.

it was shit. here’s mine:

carve your data

let me explain. i often write code like this:

def read_excel(excel):
    data = pandas.read_excel(excel)
    data = data.to_dict()
    data = [datum['x'] for datum in data if datum['y'] == 42]
    return data

i know. it’s a cardinal sin. data is a dataframe, then a dict, then a list of “stuff”.

it breaks the first commandment of:

thou shalt have variables that behave predictably

and thou shalt not reuse them

wtf art thou doing?

but compare this to the alternative.

def read_excel(excel):
    df = pandas.read_excel(excel)
    dict_data = df.to_dict()
    filtered_data = [datum['x'] for datum in dict_data if datum['y'] == 42]
    return filtered_data

there is a difference.

anyone reading the first code will understand: i want my data in a certain way, and i will not rest until i get it.

i call that carving data.

and it matters a lot. i spend a lot of time doing this. but the logic flows effortlessly afterwards.

carve it.

now.

17/06/26

okay no i changed my mind.

the most important tip is folding.

fold

an algorithm is a list of instructions.

when you start, your simple code is a simple list of instructions.

code gets bigger.

it’s still a list of instructions, those instructions just became list themselves. of instructions. it’s turtles all the way down.

and you never fold?

fold my son, fold

everywhere i look… nobody folds, people use a minimap like it’s lol, split code in a bazillion files, why, why? why suffer?!

i swear i should make a code editor called “origami” where all shortcuts are fold as rehab therapy.

done for now? fold

ctrl-shift-/: fold.

ctrl-shift-p> command panel> type “fold”> “fold all”.

and the second most important advice would be: comment. i mapped ctrl-/ to comment. look how close it is to fold shortcut.

and then yeah carve your data or whatever, who cares.