Commonroad-crime

Updated
- the saving of the results to xml files with the format of `CriMe-{scenario.scenario_id}_veh_{self.config.vehicle.ego_id}.xml
Fixed
- verbose options for `compute` functions
- matplotlib version
- fix the WTTR config saving options (align with the `verbose` option)

Updated
- Visualization of criticality curve: now the nan and inf are distinguished!! The tick for inf is automatically generated!
- Versions of `commonroad-route-planner` and `commonroad-reach`
Added
- For evaluating an interval, the starting time step should be always smaller or equal than the ending one.
- save the results to xml files
Fixed
- the memory issues for P_MC measure due to all the simulated vehicle states being stored during the evaluation for subsequent visualization, now the default mode for visualization is off.
- for all measures, check whether the time step is valid in the function `validate_update_states_log`. If no, `NaN` is returned. Similar to `compute_criticality`: only compute for the valid ego vehicle and other vehicles, otherwise `NaN` is returned
- check whether the vehicles are in the same lanelet: now the lanelet is extended by its successors and predecessors
- adds the error handling of the out-of-projection-domain for many measures, such as TTC, THW, ALongReq, ALatReq
- Initialization of the ego vehicle in the base class
- fix the pf and soi measure, when the other vehicle is not present in the scenario for some time steps
- fix the dce and ttce measure of which the initial value should be math.inf
- fix the thw and soi with the support of obstacles in circular shapes

Added
- Error handling for check in same lanelet
Fixed
- The extension of the lanelet vertices for projecting between different coordinate domains
- The compatability with the new commonroad-io version
- The floating-point precision errors for orientation computation
- The setup of the CLCS attribute of each measure can be updated correctly using the `update` function in the configuration

Added
- Add the log verbose for batch evaluation
Changed
- The lanelet vertices are now extended and smoothed before constructing the curvlinear coordinate system; therefore, some of the test results vary a bit
Fixed
- The evasive maneuver simulator: now the initial time step of the simulated vehicle doesn't have to start from 0
- The adaptive turning behaviors at intersections for the TTM simulation, i.e., the velocity needs to first satisfy the desired requirements based on the intersection's curvature
- When an empty yaml file is provided for batch evaluation, it would now use the default configuration setup.
- The bridge to compute the drivable area, now it is able to get the scenario from the same folder
- The turning simulation: the last stage orientation error

Fixed
- The original equation for computing TTC is wrong, now its fixed (also updated in the paper)
- fix the TCI, whose optimization problem was not reset each time before
- verbose for each measure and the batch evaluation
- without updating the default vehicle states when creating the other vehicle (together with the `CommonRoad-io==2023.3`)
Added
- support for Python 3.10 and preparation for 3.11
- add code style checking (black)
- sequential batch evaluation
- add the following new measures
- MSD, PSD
- CI, CPI
- ET, PET
- SOI
Updated
- update the installation of commonroad-reach using pip install, the corresponding code part is updated

Fixed
- fix the way of obtaining the default configurations, which didn't work for using the crime with the pip package
- fix the initialization of simulators with NONE maneuver