Fix issue 45

[thibautjombart]

This implements the following changes to fix https://github.com/reconhub/trendbreaker/issues/45. This should be safe to use with stratified data, varied time intervals, and non-ordered data.

• explicit definition of the time variable through date_index
• adding accessors for trendbreaker outputs (get_formula and get_response)
• fixing hard-coded response variable in plot
• new functions to define training set, now robust to stratification and non-ordering of dates in the input: get_training_data and set_training_data
• fixed plots for incidence2 inputs
• reverts to simulate_pi = TRUE by default

Passes the check on:

> R.version
               _                           
platform       x86_64-pc-linux-gnu         
arch           x86_64                      
os             linux-gnu                   
system         x86_64, linux-gnu           
status                                     
major          4                           
minor          0.3                         
year           2020                        
month          10                          
day            10                          
svn rev        79318                       
language       R                           
version.string R version 4.0.3 (2020-10-10)
nickname       Bunny-Wunnies Freak Out     

[TimTaylor]

Looks good to me. I've made a couple of minor additional commits. I'll let you check these first and then we can merge.