quantcn
quantcn
Hi, Author, thank you for your good work. I have a problem, please see below. I have two datas: 1minute and 5minute bar. But in he next, only a 5miniute...
get_feature_importance() will return importance value of column_xx not features' name. How to make it return importance value of features' name.
在例子文件workflow_by_code.ipynb生成累计收益率图表的代码是: label_df = dataset.prepare("test", col_set="label") label_df.columns = ['label'] pred_label = pd.concat([label_df, pred_df], axis=1, sort=True).reindex(label_df.index) analysis_model.model_performance_graph(pred_label) # 生成累计收益率图表 我的问题是,这里label_df 取得的标签label是预处理后(比如经过标准化处理)的推理数据集标签,并不是原始的收益率,那么据此生成累计收益率图表是不是不合适,我觉得应该采用原始label(代表未经标准化处理的原始收益率)来生成累计收益率图表。 不知我的理解对否?
您好,请问能把内置的如下处理器分个类吗:哪些是共享处理器,哪些是学习处理器,哪些是推理处理器? 并指出哪些是处理特征的哪些是处理标签的? DropnaProcessor: processor that drops N/A features. DropnaLabel: processor that drops N/A labels. TanhProcess: processor that uses tanh to process noise data. 用于特征还是标签? ProcessInf: processor that handles infinity values,...
When we use R.save_objects(trained_model=model) It generates a file "trained_model ", I wonder if this is a pkl file, same as trained_model.pkl?
很多同学希望将Qlib用于A股实盘操作,但Qlib本身不含实盘接口,无法直接连接实盘。 好在通常我们使用Qlib是基于日线数据,定期调仓再平衡。这种低频操作完全可以实现一种间接的实盘操作,那就是利用Qlib来执行预测和生成订单,然后手工下单。这样,如果是按周再平衡,则每周只需操作一次,完全可以达到实盘效果。 [这里有个视频](https://mp.weixin.qq.com/s/cl1zgQTRCTaHe-NAM-RWMg)就是讲解实盘中如何利用Qlib执行预测生成订单,并手工执行订单的流程。
hello, very interesting work you've done. I have a problem when runing your code. When runing the following code: import FinanceDataReader as fdr import datetime as dt start_date = dt.datetime(2011,1,1)...