Huang Li

I think simply using slice assignment would be ok: ```python def __call__(self, df): cols = get_group_columns(df, self.fields_group) df.loc[:, cols] = df.loc[:, cols].fillna(self.fill_value) return df ```

Or if you want to use numpy to accelerate (I can achieve 10x speed), you should assign the df.values (or df.to_numpy()) to a variable first, then fill and assign back:...

It seems like you cannot use absolute path or special characs. I met the same problem and I solve it using relative path with no special characs.

> Hi, [@LeetaH666](https://github.com/LeetaH666) Thank you for your attention to `qlib`. How can we reproduce this issue? I can not provide a one-click reproducing script, but the steps are: 1. dump...