One can manually add the key cyclones to the returned dictionary.

Replace `tests_old_dir` by PWD for tests in the same series.

Move content of `tests.json` to the end `Tests/short_tests.json`.

There were only three files in `Overlap/Analysis`.

There are other files in `Overlap/Tests/Input` with the same content as
removed files. The redundancy came from commit c58c21b7.

Add `node_id_param.json` to requirements of `report_graph.py`.

Since we now produce from Fortran edgelists with node indices, we use
`report_graph_int.py`.

Instead of filenames constructed from the orientation. So we need to
create symlinks before running the script. It is clearer this way that
the script processes a single orientation, and that the processing
does not depend on the orientation. The script does not take a command
line argument any longer.

Rename variable `N(anti|cyclo)` to `N_eddies` in file
`N(anti|cyclo).mat`. Motivation: allow more transparent processing in
`overlap_v6.py`.

Rename `edgelist_no_header.csv` to `edgelist_unsorted.csv`.

Write file `node_id_param.json` from programs `test_overlap` and `eddy_graph`.

Write node indices instead of dates and eddy indices in
edgelist. Compute `e_overestim` from `ishape_last`. Remove header
lines of edgelist. See commit 3370733a. Replace argument d of procedure
overlap by arguments `e_overestim` and k.

Do not create file `number_eddies.csv` in program `eddy_graph`. Since
there are no interpolated eddies any longer, the file
`number_eddies.csv` just contained the same information that is in
`ishape_last.txt` in the input SHPC. Remove argument
`unit_number_eddies` of procedure `dispatch_snapshot`.

Do not create attribute interpolat of nodes in graph. There are no
longer nodes that should be drawn with a box.

No longer useful since there are no interpolated eddies any longer.

Remove component `number_eddies` of derived type snapshot. It was the
same as `number_vis_extr` since there are no interpolated eddies any longer.

Thus remove too component interpolated of derived type eddy and
component `extr_interp` of derived type shpc.

In procedure overlap, write an edge across delta days instead of delta
edges across single days. So we do not need procedure `write_overlap`
any longer, nor program `test_write_overlap`.

Do not try to read attributes of interpolated eddies when reading a
graph. Remove option `--interp` of scripts `report_graph.py` and
`report_graph_int.py`. Remove argument `read_interp` of function
`read_eddy_graph` in both scripts.

As the namelist is now shorter, it is no longer convenient to write it
into a file.

Remove tests `Write_overlap_different` and `Write_overlap_shifting`:
they are no longer different from test `Write_non_successive_overlap`.

This is a big change, a big simplification. Creating interpolated
eddies was the way of Alexis Chaigneau. Rémi L. thinks it is no longer
useful.

So remove procedure `init_interpolated_eddy`; do not create SHPC for
interpolated eddies in programs `eddy_graph`, `test_overlap` and
`test_write_overlap`; remove argument `hshp_interp` of procedure
overlap; remove arguments `hshp_interp`, m, cyclone, `coord_extr_1`,
`ssh_extr_1`, `coord_extr_2` and `ssh_extr_2` of procedure
`write_overlap`.

Consequence of commit 664f62bb.