Update something

[oooo26]

• delete unused parameters
• pack specific parameters (add function: initial_setting & clear_setting)
• pack restore_for_normal
• restructure gs_path & gs support CV results output
• rewrite screening (now use algorithmXXX.fit() and unlink with model_fit.h)
• API update
• simplify algorithm pointer (use algorithm_list_xxx uniformly)
• move beta_size to Algorithm class

[Mamba413]

Have you checked the code via Pytest?

[oooo26]

Yes, it has passed Pytest.

[Mamba413]

Yes, it has passed Pytest.

Have you changed anythings in pytest?

[oooo26]

In PCA test, I find the sparse test actually give a dense matrix, so I change it to a coo_matrix there. Others are unchanged.