search

lin-tan / CURE

For our ICSE21 paper "CURE: Code-Aware Neural Machine Translation for Automatic Program Repair" by Nan Jiang, Thibaud Lutellier, and Lin Tan
https://www.cs.purdue.edu/homes/lintan/publications/cure-icse21.pdf
Other
50 stars 17 forks source link

Issues on replicating results of Defects4J #18

Open thanhlecongg opened 1 year ago

thanhlecongg commented 1 year ago

Hi,

I'm trying to replicate the results following your instructions and got similar results for QuixBugs. However, when I ran CURE for Defects4J, I got very different results.

This is an example results in reranked_patches for Chart-1: Screenshot 2023-08-02 at 5 17 18 pm

I think this may be because of the difference in the metadata of Defects4J such as vocabulary.txt. However, I don't what the is difference as the instructions for Defects4J are not provided. I'd really appreciate it if you could give me instructions to run CURE on the Defects4J dataset.

Many thanks.

zombieQQ commented 1 year ago

The same issues.Did you solve this problem?

zombieQQ commented 1 year ago

I even can't generate the patches of the testcase ,also the same issue. image @jiang719 看到的话麻烦解答一下，非常感谢