How can run the new rdagent fin_quant?

[heimo88]

❓ Questions and Help

Hi, why I changed the conf_baseline/conf_combined_factors_sota_model/conf_combined_factors.yaml the train: [2008-01-01, 2014-12-31]
valid: [2015-01-01, 2016-12-31]
test: [2017-01-01, 2020-08-01] to train: [2018-01-01, 2021-12-31]
valid: [2021-01-01, 2022-12-31]
test: [2022-01-01, 2023-12-31]
and then I run rdagent fin_quant the range still show
train:[2008-01-01, 2014-12-31]
valid: [2015-01-01, 2016-12-31]
test: [2017-01-01, 2020-08-01]
How to make the settings take effect for the rdagent fin_quant?

[heimo88]

I found that the qlib_factor_experiment_setting template in rdagent\scenarios\qlib\experiment\prompts.yaml is fixed and does not change according to the configuration file, right?

[SunsetWolf]

Hi, @heimo88 The problem you are experiencing is with the expected behavior. In the fin_quant scenario, modifying the YAML config files will not change the displayed train/valid/test ranges.

RD-Agent now supports customizing time segments via the .env file instead. Please configure the train/valid/test dates there to make them take effect.

You can find the detailed instructions here.

Also, you are correct that the qlib_factor_experiment_setting template in prompts.yaml is fixed and does not dynamically reflect YAML config changes.