Remove DataFrame.append from the 10min intro

[TomAugspurger]

We show it https://pandas.pydata.org/pandas-docs/stable/getting_started/10min.html#append.

I don't think we should recommend it there, especially without providing context that the complexity of DataFrame.append is very different from list.append.

[sameshl]

@TomAugspurger I understand what you mean. But don't you think that for a new user append function might be useful?

[TomAugspurger]

No I don't think so. I think for that use case we should recommend building up a list of rows, then wrapping in a Series / DataFrame, then appending. I'm not sure whether something like that belongs in a 10-minute introduction though.

…

On Mon, Jul 22, 2019 at 9:18 AM Samesh Lakhotia ***@***.***> wrote: @TomAugspurger <https://github.com/TomAugspurger> I understand what you mean. But don't you think that for a new user append function might be useful? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#27518?email_source=notifications&email_token=AAKAOIS54X75GJP47NZBMDTQAW6SZA5CNFSM4IFY4IDKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2QCD5Y#issuecomment-513810935>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAKAOIQEG3G76PFPTDEILDTQAW6SZANCNFSM4IFY4IDA> .

[sameshl]

Yeah, you are right. dataframe.append is not as trivial as list.append. I could fix this issue then

[ParfaitG]

Agreed with @TomAugspurger . Occasionally on StackOverflow, you see new users running DataFrame.append inside a loop in same fashion as they would with list.append , leading to the hazardous quadratic copy. Therefore, leaving it out of an intro to new users may be beneficial:

# INITIALIZE AN EMPTY DATA FRAME
df = pd.DataFrame()

for i in range(5):
    # GROW DATA FRAME IN LOOP
    df = pd.DataFrame({'int': np.random.randint(1, 10, 500),
                       'num': np.random.uniform(0, 1, 500)
		       })
    df = df.append(df)

As the numpy/pandas guru, @unutbu admonishes: Never call DataFrame.append or pd.concat inside a for-loop. It leads to quadratic copying.