Conversation
|
| qlib_init: | ||
| provider_uri: "~/.qlib/qlib_data/yahoo_cn_1min" | ||
| region: cn | ||
| market: &market ['SH605222', 'SZ002796', 'SZ002246', 'SZ000713', 'SZ002820', 'SH601328', 'SZ000668', 'SH603359', 'SZ002144', 'SH600195', 'SH603685', 'SH603386', 'SZ002586', 'SZ000573', 'SZ000605', 'SZ002842', 'SH600068', 'SZ300547', 'SZ000926', 'SZ002036', 'SZ002161', 'SH600715', 'SZ300427', 'SZ002573', 'SZ300142', 'SH605116', 'SZ002951', 'SH600276', 'SZ002437', 'SH603355', 'SZ002893', 'SH600584'] |
There was a problem hiding this comment.
Waiting for @zhupr to update the instruments list
There was a problem hiding this comment.
updated: python get_data.py qlib_data --name qlib_data --target_dir ~/.qlib/qlib_data/cn_data_1min --interval 1min --region cn
qlib/contrib/eva/alpha.py
Outdated
| def calc_prec( | ||
| pred: pd.Series, label: pd.Series, date_col="datetime", quantile: float = 0.2, dropna=False, is_alpha=False | ||
| ) -> Tuple[pd.Series, pd.Series]: | ||
| """calculate the precision |
There was a problem hiding this comment.
Will excess return/return a more comprehensible name than precision?
qlib/contrib/eva/alpha.py
Outdated
| long precision and short precision in time level | ||
| """ | ||
| if is_alpha: | ||
| label = label - label.mean(level=0) |
There was a problem hiding this comment.
You should use groupby('datetime')
There was a problem hiding this comment.
label.mean(level=date_col)
| self.params.update(kwargs) | ||
| self.model = None | ||
|
|
||
| def _cal_signal_metrics(self, y_test, l_cut, r_cut): |
There was a problem hiding this comment.
Why don't you use calc_prec here?
qlib/workflow/record_temp.py
Outdated
| self.get_path("long_pre.pkl"), | ||
| self.get_path("short_pre.pkl"), | ||
| ] | ||
| paths.extend([self.get_path("long_short_r.pkl"), self.get_path("long_avg_r.pkl")]) |
There was a problem hiding this comment.
Why is extend a better choice than writing it in the list directly here?
|
@javaThonc Please remember to update the evaluation results on high frequency data and send us the profiling results. Besides, don't forget to update the results of tabnet . |
update high freq demo
* spaceship: format type of y = pd.series; fix a wrong spelling in xgb * s3e11: format of y -- pd.series * spaceship: format of y & fit nn * spaceship: wrong spelling in xgb * ci issue
3 participants
Description
Create a high-frequency demo with the calculation of precision to calculate the right prediction level and a classification version of lightGBM that is designed for the high-frequency data classification method as well as a workflow that can be used to backtest later.
Motivation and Context
Create a high freq demo that can be used to test whether a dataset contains signal for trading
Types of changes