data_to_distribution (and parameter_to_distribution) function improvements

[SeniorKate]

1. Implement date arguments to specify the pairs of dates

e.g. instead of: day_diff <- data$notification_date - data$specimen_date

have day_diff <- date1 - date2

Where date1 and date 2 are arguments to specify from your data

2. Have an example of setting min_delay with a quantile from the data (i.e. changing from the default of 0 to 0.1 etc)