I added a graph to compare the k estimations according to the "level of contact" between the inoculated birds and the inidrect contact birds. Finally, this graph won't be used because but I let it just in case.

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 20 ou 21

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 10 ou 14

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 20 ou 21

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 20 ou 21

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 12 ou 14

2/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

3/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 11 ou 12

2/ We export the data_sir (for each trial, if needed) in data_sir_export folder.

3/ The function calcul_t_end_dt_L is used to identify the time on which L is calculated, and the dt associated with these times. Finally, this function has been simplified because all the experiments consider time from t0 to tmax-dt + dt is unchanged (last dt is already associated with tmax-dt time). But, I left the function.

4/ I made a common introduction for parameters estimations, and then details for Method 1/ and 2/ which change the IC estimations.

=> Same estimations than before all theses changes + estimations are also the same for t_end (last t for L calcul) = 72 ou 96

There was a mistake: at time=0 there was no virological sample, so the result is na and no 0 or neg (and na will be considered as 0 by our function).

We created functions to be used in each Rmd to: 1/ identify the t used for L calcul (even if finally it is always until tmax-dt) and 2/ to executes optim function in several experimental conditions.

The data frame sir is a bit modified because of the function data_sir_C = the definition of C was different in the previous version, but the estimation changes of k and beta_within are very little.

Before, the delay_I_D was equal to 2, whereas it is estimated at 1 in this version with the info of all the groups and all the trials.

What's more, the false negative were not taken into account, but they are in this version. For instance in the trial 4, i2(14) = 6 (instead of 5 before). It is 6 because the penultimate bird  (see individual raw data) has a false negativ at day 14.

As before, there is no i2(1) (thanks to data_raw_first_pos_ind function, but it has already been done in the previous version).

k and beta_within estimations have a little changed.

Now we have i2(2) = 0, C2(2) = 4 and i2(4) = 4.Finally, thanks to the data_raw_first_pos_coll function, we do not have first CD or CI infectious at the same time that first inoculated infectious. In fact, before it was impossible to have a new case at time 1 whereas prevalence was equal to 0.

It changes just a bit the estimations of k and beta_within.

So, in the new version we have S2(9) = 1.

However, this bird becomes infectious at day 10, and the case C is only recorded at day 8 (as before), because prevalence is equal to 0 at day 9. So, the estimates of k and beta_within are not different, althought the data frame sir is a bit changed.

No change in k and beta_within estimation after the changes of october & november 2024 to harmonize the functions.

No change in k and beta_within estimation after the changes of october & november 2024 to harmonize the functions.

No change in k and beta_within estimation after the changes of october & november 2024 to harmonize the functions.

No change in k and beta_within estimation after the changes of october & november 2024 to harmonize the functions.

Raw data have been harmonized to always have the same structure (according to individual raw data or collective raw data).

--> Individual results:
rows --> as many as individuals
columns --> id
        --> group = 'inoc', 'CD', 'CI'
        --> trial
        --> day_min ... day_max or hour_min ... hour_max
data --> 'pos', 'neg', 'dead', 'na', '0' (not in the experiment yet)

--> Collective results:
rows --> n(sample time) * n(group) * n(trial)
columns --> day or hour
        --> group = 'inoc', 'CD', 'CI'
        --> trial
        --> pos
        --> total

Parameter estimatation was harmonised across all experiments. Two types of articles were identified, according to the two types of raw data = individual (individual result by bird = pos, neg, dead, or 0 = not yet introduced into the experiment) or collective (we only have the number of infectious birds over time and the number of live birds over time).

These functions are also detailed in the power point available in the folder, or at the following link: https://docs.google.com/presentation/d/1B90eFxqEGJzvxgoyjYjbZ8KY8qgmQ-7_0FBzR0jWNmA/edit#slide=id.g31422fecb95_2_75.

It is also available on: https://docs.google.com/presentation/d/1B90eFxqEGJzvxgoyjYjbZ8KY8qgmQ-7_0FBzR0jWNmA/edit#slide=id.g31422fecb95_2_75