data frame and vector min and max slow

Hi, using `.min` and `.max` for vector is too slow for large dataset (3.5m records) than using the `.to_a.min` and `.to_a.max`.

```
vector.describe:          99.064083   0.171811  99.235894 ( 99.350809)
vector.to_a.min:           0.056799   0.000000   0.056799 (  0.056925)
vector.min:                39.123274   0.087998  39.211272 ( 39.260956)
```

Same goes for `percentile` and `median`.


