HighD dataset #37
HighD dataset #37
Conversation
… to upstream master
|
My merging commit is weird... I had to edit manually stuff that was not supposed to be there. Maybe a GitHub bug? |
| 'Right Alongside ID', | ||
| 'Right Following ID', | ||
| 'Lane Identification' | ||
| ), |
There was a problem hiding this comment.
Haha, I think you could have used an ordered dict, and then you could have simply dumped the keys 😜
There was a problem hiding this comment.
Good point. Do you want me to change this, or should we leave it for now?
There was a problem hiding this comment.
If you want to change it, I can wait before merging.
|
|
||
| car_sizes = dict() | ||
| for ts in time_slots: | ||
| d = df[ts] | ||
| car = lambda i: d[d['Vehicle ID'] == i] | ||
| def car(i): return d[d['Vehicle ID'] == i] |
There was a problem hiding this comment.
PEP8's fault not mine...
| @@ -47,4 +109,5 @@ | |||
| car_sizes[ts][c] = size | |||
| print(c) | |||
|
|
|||
| torch.save(car_sizes, 'traffic-data/state-action-cost/data_{}_v0/car_sizes.pth'.format(opt.map)) | |||
| torch.save(car_sizes, f'traffic-data/state-action-cost/data_{opt.map}_v0/car_sizes.pth') | |||
| parser.add_argument('-delta_t', type=float, default=0.1) | ||
| parser.add_argument('-recording', type=str, default="01", |
There was a problem hiding this comment.
Can't we use time_slot instead? They are used for the exact same purpose, no?
There was a problem hiding this comment.
Good point. I think this is an artifact of me trying to keep this new script as similar to map_i80 as possible
There was a problem hiding this comment.
Let's use time_slot for highd too.
| @@ -43,6 +44,10 @@ | |||
| delta_t=opt.delta_t, | |||
| ) | |||
|
|
|||
| # HighD dataset will use recordings IDs rather than time slots | |||
| if opt.map == 'highD': | |||
| kwargs['rec'] = opt.recording | |||
| #SBATCH --mail-type=FAIL # notifications for job fail | ||
| #SBATCH [email protected] | ||
|
|
||
| source activate PPUU |
There was a problem hiding this comment.
This is not supposed to work... unless there's this before:
eval "$(conda shell.bash hook)"
Enable code base to incorporate highD dataset into project pipeline
New scripts and files:
map_highD.py: Main script that creates environment and car classes for highD datasetscripts/run_generate_data_highD.shscripts/submit_generate_data_highD.slurmModified files:
play_maps.py: Enable playback of highD mapextract_car_size.pyandgenerate_trajectories.py: Enable data processing of highDdataloader.py: Enable batch loading of highD