Alter inference ends file to match structure of prediction full file

[peter-t-fox]

The prediction full file now has the following format, where the epidemiological compartments are all located on a single row.

iter, day, age_group, S, E, E_t, I_p, I_t, I1, I2, I3, I4, I_s1, I_s2, I_s3, I_s4, H, R, D
0, 0, 0, 1266783, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
...

The full file produced in inference mode is formatted as:

iterID,age_group,comparts,value
2, 0, 0, 780556
2, 0, 1, 111340
2, 0, 2, 0
2, 0, 3, 39191
2, 0, 4, 0
2, 0, 5, 56667
2, 0, 6, 48062
2, 0, 7, 39969
...

where each compartment gets a row to itself.

Inference I/O code should be updated to match the prediction mode format.

[github-actions[bot]]

Heads up @thibaud-porphyre @peter-t-fox - the "Covid19_EERAModel" label was applied to this issue.

[kzscisoft]

• [X] Change headers to match as above
• [x] Produce all output files in an identical manner using similar WriteInferenceXHeader functions etc