How to use _scenario method in Molotov

Best Python code snippet using molotov_python

experiment_runner.py

Source:experiment_runner.py

...105    polygon_9_gen = PolygonGenerator(radius, 9)106    107    scenario_id = 1108    109    def _scenario(**kwargs) -> TestScenarioSpec:110        nonlocal scenario_id111        scenario = TestScenarioSpec(112            id=scenario_id, common=common_spec, **kwargs113        )114        scenario_id += 1115        return scenario116    117    std_shape = dict(shape_gen=rectangle_gen, n_groups=6)118    all_affine = dict(use_translation=True, use_rotation=True, use_scale=True)119    120    scenarios = (121        _scenario(**std_shape),122        123        _scenario(use_translation=True, **std_shape),124        _scenario(use_rotation=True, **std_shape),125        _scenario(use_scale=True, **std_shape),126        127        _scenario(use_noise=False, **std_shape, **all_affine),128        _scenario(use_noise=True, **std_shape, **all_affine),129        130        _scenario(131            use_noise=True, shape_gen=polygon_5_gen, n_groups=6, **all_affine132        ),133        _scenario(134            use_noise=True, shape_gen=polygon_7_gen, n_groups=6, **all_affine135        ),136        _scenario(137            use_noise=True, shape_gen=polygon_9_gen, n_groups=6, **all_affine138        ),139140        _scenario(141            use_noise=True, shape_gen=rectangle_gen, n_groups=3, **all_affine142        ),143        _scenario(144            use_noise=True, shape_gen=rectangle_gen, n_groups=5, **all_affine145        ),146        _scenario(147            use_noise=True, shape_gen=rectangle_gen, n_groups=7, **all_affine148        ),149        _scenario(150            use_noise=True, shape_gen=rectangle_gen, n_groups=9, **all_affine151        ),152    )153    154    run_experiments(155        scenarios, config.RESULTS_FILE_PATH, config.N_INSTANCES_PER_SCENARIO156    )157    158    return 0159160161if __name__ == '__main__':
...

plot_fitCanvas.py

Source:plot_fitCanvas.py

1# script that prints all fit_canvas that can be found in all the passed root files and stores them with2# hopefully reasonable names3#4# TODO: this works quite nicely for the vtx and eta scenario already, however the pt_abseta scenario is5# currently a bit of a mess at the moment.6import os7import subprocess8import glob9import re10def saveCanvasOfFile(_file, _regex="fit_canvas", _extension="pdf"):11    """12    Saves all the TCanvas found in the file matching the passed regex via a call to the saveCanvas exectuable13    """14    path_to_exe="/afs/hephy.at/work/t/tmadlener/CMSSW_8_0_12/src/TnPfromJPsi/PlotEfficiency/utils/saveCanvas"15    DEVNULL = open(os.devnull, 'w') # needed to dump the output of TCanvas::Save into /dev/null16    print("Saving TCanvas matching \'{}\' from file \'{}\'".format(_regex, _file))17    status = subprocess.call([path_to_exe, _file, _regex, _extension], stdout=DEVNULL, stderr=subprocess.STDOUT)18    return status19def commonFileNameCleanup(_filename, _ID, _scenario, _trigRgx):20    """21    Do the cleanup that is common to both mvAndRename functions below:22    Remove trigger stuff and the TDirectory information from the pdf filename23    """24    fn = _trigRgx.sub('', _filename)25    fn = fn.replace(":tpTree:", "").replace("{}_{}:".format(_ID, _scenario), "")26    fn = fn.replace("_pair_drM1_bin0_", "").replace("_pair_probeMultiplicity_bin0_", "")27    return fn28def mvAndRename(_ID, _scenario, _targetdir="fitCanvasPdfs", _extension="pdf"):29    """30    The output of saveCanvas is not that nice and has to be cleaned up.31    Mainly doing some replacing of the trigger parts and moving the created files to a separate directory32    """33    triggerRegex = re.compile('(_tag)?_Mu7p5_Track2_Jpsi(_(TK|MU)_pass_)?')34    for filename in glob.glob(":tpTree:{}_{}*:*.{}".format(_ID, _scenario, _extension)):35        filenameTo = commonFileNameCleanup(filename, _ID, _scenario, triggerRegex)36        dirTo = "{}/{}_{}".format(_targetdir, _ID, _scenario)37        if not os.path.exists(dirTo):38            os.makedirs(dirTo)39        filenameTo = "/".join([dirTo, filenameTo])40        os.rename(filename, filenameTo)41def mvAndRenamePt(_ID, _scenario, _file, _targetdir="fitCanvasPdfs", _extension="pdf"):42    """43    The output of saveCanvas is not that nice and has to be cleaned up.44    Mainly doing some replacing of the trigger parts and moving the created files to a separate directory.45    For pt we currently run into a memory problem and have to split the input into abseta different bins.46    However this is not reflected in the names of the produced pdfs (since the .root files do not "know"47    about this splitting, so we have to define a special version for renaming the pt scenario that uses48    some more information that can be obtained from the root file (_file)49    """50    # compute some additional info from the filename (that follows at least for the moment a common pattern)51    addInfo = _file.replace(".root", "")52    addInfo = addInfo.replace("{}_{}_".format(_ID, _scenario), "")53    addInfo = addInfo.replace("TnP_MuonID_","").replace("_data_all_", "").replace("_signal_mc_", "")54    triggerRegex = re.compile('(_tag)?_Mu7p5_Track2_Jpsi(_(TK|MU)_pass_)?')55    for filename in glob.glob(":tpTree:{}_{}*:*.{}".format(_ID, _scenario, _extension)):56        filenameTo = commonFileNameCleanup(filename, _ID, _scenario, triggerRegex)57        dirTo = "{}/{}_{}_{}".format(_targetdir, _ID, _scenario, addInfo)58        if not os.path.exists(dirTo):59            os.makedirs(dirTo)60        filenameTo = "/".join([dirTo, filenameTo])61        os.rename(filename, filenameTo)62def processAllFiles(_dir, _ID, _scenario,63                    _targetdir="fitCanvasPdfs",64                    _canvasRegex="fit_canvas",65                    _extension="pdf"):66    """67    Process all .root files matching the _ID AND _scenario in _dir.68    """69    os.chdir(_dir)70    for f in glob.glob("TnP_MuonID_*_{0}_{1}*.root".format(_ID, _scenario)):71        saveCanvasOfFile(f, _canvasRegex, _extension)72        if "pt_abseta" in _scenario:73            mvAndRenamePt(_ID, _scenario, f, _targetdir, _extension)74        else:75            mvAndRename(_ID, _scenario, _targetdir, _extension)76# Define on what to run77IDs = ["Loose2016", "Medium2016", "Tight2016", "Soft2016"]78scenarios = ["eta", "vtx", "pt_abseta"]79basedir="/afs/hephy.at/work/t/tmadlener/CMSSW_8_0_12/src/data_rootfiles"80targetdir="/afs/hephy.at/work/t/tmadlener/CMSSW_8_0_12/src/outputfiles/Figures/FitCanvasPdfs/data"81# pdf, ps and svg are currently working, png should too according to the ROOT documentation but doesn't at the moment82# Workaround to get png is to save the plots as ps and convert them to png using gs83plotformat="pdf"84for ID in IDs:85    for scen in scenarios:86        print("Currently processing ID: {}, scenario: {}".format(ID, scen))...

test_formatter_rerun.py

Source:test_formatter_rerun.py

...11    formatter_name = "rerun"12    def test_feature_with_two_passing_scenarios(self):13        p = self._formatter(_tf(), self.config)14        f = self._feature()15        scenarios = [ self._scenario(), self._scenario() ]16        for scenario in scenarios:17            f.add_scenario(scenario)18        # -- FORMATTER CALLBACKS:19        p.feature(f)20        for scenario in f.scenarios:21            p.scenario(scenario)22            assert scenario.status == "passed"23        p.eof()24        eq_([], p.failed_scenarios)25        # -- EMIT REPORT:26        p.close()27    def test_feature_with_one_passing_one_failing_scenario(self):28        p = self._formatter(_tf(), self.config)29        f = self._feature()30        passing_scenario = self._scenario()31        failing_scenario = self._scenario()32        failing_scenario.steps.append(self._step())33        scenarios = [ passing_scenario, failing_scenario ]34        for scenario in scenarios:35            f.add_scenario(scenario)36        # -- FORMATTER CALLBACKS:37        p.feature(f)38        for scenario in f.scenarios:39            p.scenario(scenario)40        failing_scenario.steps[0].status = "failed"41        assert scenarios[0].status == "passed"42        assert scenarios[1].status == "failed"43        p.eof()44        eq_([ failing_scenario ], p.failed_scenarios)45        # -- EMIT REPORT:46        p.close()47    def test_feature_with_one_passing_two_failing_scenario(self):48        p = self._formatter(_tf(), self.config)49        f = self._feature()50        passing_scenario = self._scenario()51        failing_scenario1 = self._scenario()52        failing_scenario1.steps.append(self._step())53        failing_scenario2 = self._scenario()54        failing_scenario2.steps.append(self._step())55        scenarios = [ failing_scenario1, passing_scenario, failing_scenario2 ]56        for scenario in scenarios:57            f.add_scenario(scenario)58        # -- FORMATTER CALLBACKS:59        p.feature(f)60        for scenario in f.scenarios:61            p.scenario(scenario)62        failing_scenario1.steps[0].status = "failed"63        failing_scenario2.steps[0].status = "failed"64        assert scenarios[0].status == "failed"65        assert scenarios[1].status == "passed"66        assert scenarios[2].status == "failed"67        p.eof()68        eq_([ failing_scenario1, failing_scenario2 ], p.failed_scenarios)69        # -- EMIT REPORT:70        p.close()71class TestRerunAndPrettyFormatters(MultipleFormattersTest):...

multi_agent_race.py

Source:multi_agent_race.py

1import math2from typing import Dict, Union3import gym4from .scenarios import MultiAgentScenario5from ..core.definitions import Pose, Velocity6class MultiAgentRaceEnv(gym.Env):7    metadata = {'render.modes': ['follow', 'birds_eye', 'lidar']}8    def __init__(self, scenario: MultiAgentScenario):9        self._scenario = scenario10        self._initialized = False11        self._time = 0.012        self.observation_space = gym.spaces.Dict([(k, a.observation_space) for k, a in scenario.agents.items()])13        self.action_space = gym.spaces.Dict([(k, a.action_space) for k, a in scenario.agents.items()])14    @property15    def scenario(self):16        return self._scenario17    def step(self, action: Dict):18        assert self._initialized, 'Reset before calling step'19        observations = {}20        dones = {}21        rewards = {}22        state = self._scenario.world.state()23        for id, agent in self._scenario.agents.items():24            observation, info = agent.step(action=action[id])25            state[id]['observations'] = observation26            done = agent.done(state)27            reward = agent.reward(state, action[id])28            observations[id] = observation29            dones[id] = done30            rewards[id] = reward31        self._time = self._scenario.world.update()32        return observations, rewards, dones, state33    def set_state(self, agent: str, pose: Pose):34        self._scenario.agents[agent].reset(pose=pose)35    def reset(self, mode: str = 'grid'):36        if not self._initialized:37            self._scenario.world.init()38            self._initialized = True39        observations = {}40        for agent in self._scenario.agents.values():41            obs = agent.reset(self._scenario.world.get_starting_position(agent=agent, mode=mode))42            observations[agent.id] = obs43            self._scenario.world.reset()44            self._scenario.world.update()45        return observations46    def render(self, mode='follow', agent: str = None, **kwargs):47        if mode not in MultiAgentRaceEnv.metadata['render.modes']:48            raise NotImplementedError(f'"{mode}" is no supported render mode. Available render modes: {MultiAgentRaceEnv.metadata["render.modes"]}')49        if agent is None:50            agent = list(self._scenario.agents.keys())[0]...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.