onlyzdd / ecg-diagnosis Goto Github PK

dxs = [dx_dict.get(code, '') for code in row['dx'].split(',')]

AttributeError: 'int' object has no attribute 'split'

Any help will be appreciate. Thanks

请问这个数据集是原始心电数据吗

您好楼主，我在执行shap_value.py这个文件时，当运行到y.append(result[:,i].sum())这一步时出现错误，提示too many indices for array: array is 1-dimensional, but 2 were indexed，请问您有出现过这个原因吗

您好，麻烦问下在运行python baselines.py --data-dir data/CPSC --classifier LR的时候，您遇到过Not enough beats to compute heart rate.这样的异常么

执行baseline.py时出现下述问题，不知楼主可否解答一下。

环境按照readme里面的重新配置的

File "baselines.py", line 53, in
df_X = generate_features_csv(features_csv, data_dir, patient_ids)
File "baselines.py", line 35, in generate_features_csv
ecg_features.append(extract_features(ecg_data))
File "/pdpd_workspace/workspace/test/ecg-diagnosis-master/expert_features.py", line 62, in extract_features
all_features += extract_channel_features(channel)
File "/pdpd_workspace/workspace/test/ecg-diagnosis-master/expert_features.py", line 54, in extract_channel_features
return stats_features + wavelet_features + hos_features
ValueError: operands could not be broadcast together with shapes (7,) (938,)

https://github.com/onlyzdd/ecg-diagnosis/blob/dfa9033d5ae7be135db63ff567e66fdb2b86d76d/main.py#L45C1-L47C23

Hi, I have the question about the above code in main.py. What is the role of labels_list and output_list here? I know in evaluate function, these two variables is to compute the F1 score. But in the train function, it seems that these two variables are not needed.

Another question is why the scheduler.step() is commented here. is this a little mistake?

Looking forward to your answer, thank you!

Hello.... I am a new graduate student interested in your work. I was trying to duplicate the results you have on your paper. I ran the main.py following your Readme file, but the resulting F1 scores seem very low. I am curious as to how I should set the parameters or arguments to have the same results with yours. Thank you.

作者您好，我在您发表的论文《Interpretable deep learning for automatic diagnosis of 12-lead electrocardiogram》补充材料中的Training and Evaluation部分看到您提到应用了十折交叉验证的方法（一共进行十轮实验，每轮都将8fold用于训练，1fold用来验证，1fold用于测试，最后取10轮的平均值作为结果），但是我在代码中没有看到相应的实现，不知道是不是我理解错了（我所看到的是只进行了一轮划分），希望能得到您的回复，万分感谢。

Hey,
When I am done training the DNN model, what should I do in order to fit the code for new testing data?
I want to use the trained model only for testing- without validation and traninig folds.

Thanks,
Nadav.

您好，我在执行完Preprocessing，Baselines，Deep model这三步的指令之后，执行Interpretation这一步的指令，出现了'No such file or directory: 'shap/shap1-CPSC\A0001.png''的错误信息。我的shao包的版本安装的是0.38.1的，请问可能是什么问题导致了这个错误呢？