Some questions about time format and dataframe

[Protomolekule]

I will start with the simple questions:

** What is the perfect dataframe for finplot?**
I've seen examples using datetime index and datetime column. Please show me df.info() and some rows.
Does the df DateTime index need to be offset-naive or offset-aware?

How to handle TimeZone and localisation?
Do I have to remove localisation before plotting?

How to set DateTime format for the crosshair info?

The crosshair is build in init:
def _x2local_t(datasrc, x): if display_timezone == None: return _x2utc(datasrc, x) return _x2t(datasrc, x, lambda t: _millisecond_tz_wrap(datetime.fromtimestamp(t/1e9, tz=display_timezone).strftime(timestamp_format)))
The timestamp_format is set also in the init:
timestamp_format = '%Y-%m-%d %H:%M:%S.%f'

but in the crosshair only up to the hrs will be displayed (in a extra row without :00). I want to have it like '29.01.24\n10:24' No idea how this will build.

The solution mentioned in the finplot/examples/snp500.py is without manipulating the time format.

Thanks for your attention

[highfestiva]

Either offset naive or aware is fine. If you wish to change the display timezone to something other than that of your computer's, you should set finplot.display_timezone. See separate questions, wiki, snippets, etc. on that.

Try setting the finplot.timestamp_format to '%d.%m.%y\n%H:%M' and see what happens.

[Protomolekule]

Hm, nothing special. I've played around with different combinations, but the minutes of finplot.timestamp_format are ignored...

Regarding Timezone: Depending on provider there are different time formats depending on candlesize etc. So I run into problems using their data. So, I came over to the idea to implement a function which normalizes the incoming datetimes to one format I will use on every chart, calculation etc.
Somtimes I see timeformat in finplot as integer and datetime with timezone etc. Therefore I'm asking about the prefered datetime format.