Add support for DataFrames and Numpy Arrays for log_plot()

[mnrozhkov]

This PR is a part of this issue: https://github.com/iterative/dvclive/issues/750

Changes:

• add support for Add support for DataFrames and Numpy Arrays in log_plot()
• implemented a new function for convert datapoints to List[Dict] dvclive/utils.py/convert_datapoints_to_list_of_dicts
• added new Python dependencies: pandas and numpy
• added a new argument to the log_plot(): columns: Optional[List[str]] = None, - optional list of column names for data in Numpy Array format

JN to test new feature: log_plots.ipynb

TODO

• [x] ❗ I have followed the Contributing to DVCLive guide.
• [x] 📖 If this PR requires documentation updates, I have created a separate PR: https://github.com/iterative/dvc.org/pull/5055

[mnrozhkov]

Thanks @mnrozhkov! Very clean and simple. I have just two overall suggestions:

1. Try to require numpy and pandas as narrowly as possible
2. Drop columns since I think users can achieve the same without it

Thanks @dberenbaum! I resolved all comments. Pls, let me know if there are other places to improve 🙌

[dberenbaum]

Thanks @mnrozhkov! Left one last comment and then I think it's ready to merge. Do you plan to also do the docs PR?

[mnrozhkov]

Thanks @mnrozhkov! Left one last comment and then I think it's ready to merge. Do you plan to also do the docs PR?

Yes, I plan to update the docs as well.