confusions when I try to recurrent the results - Githubissues

jsxlei / SCALE

Single-cell ATAC-seq analysis via Latent feature Extraction

MIT License

97 stars 17 forks source link

confusions when I try to recurrent the results #10

Open qi825 opened 4 years ago

qi825 commented 4 years ago

Dear author, I’m writing to you because I have some confusions when I try to recurrent the results using the forebrain data.I used the forebrain data (Downloaded at https://cloud.tsinghua.edu.cn/d/eb4371c556bc46ef8516/?p=%2FForebrain&mode=list),and the following were my parameters:

n_centroids=8
- lr = 0.002
- batch_size=32
- seed=18
- encode_dim=[128]
- decode_dim=[]
- latent=50
- max_iter=30000
- weight_decay=5e-4
- transform = MaxAbsScaler() And the matrix is converted to binary with cell number and peak number as 2088 and 11285. In my result ,NMI is 0.6798,and ARI is 0.5829, the NMI that you given in your paper was 0.733,and the ARI was 0.676 (E128_L50). So,could you give me some suggestions that why I can’t recurrent the results in your paper? Did I use the wrong parameters ? And what’s your parameters? I’ll be appreciated if you can help me.Thank you very much! @jsxlei

jsxlei commented 4 years ago

The lr is 0.0002 for the Forebrain dataset, you can refer to the Supplementary of our paper for the detail. The results may not be the exact same since the current version has been updated a little bit.