Request for more information about Variable Similarity Search #4
Description
Dear authors,
Recently, I read your paper about VarCLR, which is a fantastic work and gives me a lot of inspiration.
However, I have encountered some problems when replicating the experiment of Variable Similarity Search. In my attempt to construct VarSim, I followed the procedures in the paper and reproduced the variable collection from the validation set of the js150 dataset, which resulted in 550,060 variables (as opposed to 208,434 in the paper). The model's performance was also unsatisfactory on this larger dataset. Could you provide more details about VarSim construction? If possible, could you please send me a copy of the data set containing 208,434 variables? I would be extremely grateful for your help.
Thanks!