Create loglog plot using prediction from Cox model

[davidsantiagoquevedo]

This PR resolves issue #84 by creating a log-log plot using predictions from the Cox model.

Main changes

1. Added the cox_model_prediction function to calculate predictions.
2. Modified coh_eff_hr to return predictions.
3. Updated plot_loglog to generate the plot based on cox_model_prediction.
4. Updated tests to reflect these changes.

This PR includes changes from #93 and #95. The changes specific to this PR start with commit 123d5c4.

Concerning the following comments,

Is it normal in this domain to return the results for multiple models at once? I am not used to this as it can create selective reporting. I would expect the function call to specify the model.
(#76 (comment))

I'm not sure if there's a specific reason why the package runs both models simultaneously (It might be useful, but my lack of expertise in the topic can blur this). Wouldn't it be more appropriate to include a parameter for each model, or allow users to run them separately? Will all users need to run both tests?
#76 (comment)

We decided to keep returning both models to give the whole context to the estimation, since the Kaplan-Meier is not used to estimate VE but to construct the Survival Curves.