set tagID instead of master-SNAPSHOT #918
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ghost
assigned 
sfwatergit
approved these changes
Nov 12, 2018
There was a problem hiding this comment.
Reviewed 1 of 1 files at r1.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on @colinsheppard)
colinsheppard
added a commit
that referenced
this pull request
Dec 5, 2018
* Fix attributes in AOI (#881) * Subsidies working * - Move AoI to separate class - Initialize everything in one place - Ensure that access to attributes is only through AoI and not custom attributes * Fix mode subsidy to Source.fromFile independent of system * Do not read modesubsidies every time from file! * Ignore lccm spec * Ignore LCCMspec (for real this time) * Fix AoI-related broken tests * Update population in tests with attributes as necessary. * Logback to error again * Fix warmstart-routing * Ignore strange error in PA spec for now * Integrate time-dependent tolls (#875) * Un-reformat so that you can see some structure again * Test that the router reports tolls * Put linkId-based tolls into the TollCalculator; test them * Put the tuning parameter back in and test it * Router works fine without defensively waiting on it Apparently not necessary anymore * Extract Range type from ModeSubsidy; extract test into a Spec * Work on RangeSpec * Remove unused code from TollCalculator * Remove more unused code from TollCalculator * Time-dependent tolls * Tolls are used in routing * Put the real TollCalculator into the SfLight tests * Use the tuning factor also for the tolls used in routing * Ensure walking remains free * Adapt toll-prices.csv to new format * Put in a default value of travel time savings of 10$/h * Casting exception in TollCalculator * reverting... this doesn't seem to work on windows * Set akka version to `2.5.17` (#878) * Calibration docs (#877) * Calibration docs * Remove recent return (#885) * Remove return * Remove possible return * Bl/#836 unifying name scheme output files 4ci (#871) * rename all output graph images to follow the camel case * rename all output graph images to follow the camel case * change BeamGraphComparator to group the files based on new naming convention * bug-fix: 0 missing from file name in event file spec * revertning back rename of scorestats file name * Merge branch 'master' into bl/#836-unifying-name-scheme-output-files-4ci * using GUAVA library to capitalize the file names lower case in java * removed commented code part * improvised code to group file names by predefined chart prefixes * Merge branch 'master' into bl/#836-unifying-name-scheme-output-files-4ci * fixing compile time issues * renaming leghistogram files to follow standard naming convention of camel case * renaming experience_plans gz files to follow standard naming convention of camel case * rename reference_modeChoice to match standard naming convention of camel case * minor changes to file rename in beamsim * new method in beam sim to rename output files to follow the standard naming convention of camel case * Merge branch 'master' into bl/#836-unifying-name-scheme-output-files-4ci * metrics printer (#874) * feedback from carlos * feedback from Justin * convert frequencies into trips for transit routes that are specified … (#888) * convert frequencies into trips for transit routes that are specified that way * give access to NetCoordinator for third party to modify transit frequencies before doing conversion to trips * need to flip a flag to get router to use all transit trips and added freq based bus * this test no longer valid now that toll used in routing, though a fix is in the works but for now ignore * need to do freq conversion anywhere network loaded * Remove TollPriceSpec (#886) * Remove TollPriceSpec; put person-specific VOT in main routing request * Fetch person attributes locally * Add plan to person to fix test * Hopefully fix a hopefully spurious exception in a test (mocking not necessary here) * Revert "need to flip a flag to get router to use all transit trips and added freq based bus" This reverts commit c278de2. * flip flag * start work on pooling * get score test to work reliably * beam utilities (#873) * add beam-utility project * insert beam-utilities dependency, split personVehicleTransitionStats, remove jai_codec,jai_imageio * just added a line to start build again * move classes from beam to beam-utility * Improve encapsulation and avoid NPE (#876) * Improve encapsulation and avoid NPE * formatting issue * reference error * using alias for java.util.List * Spacing changes * fixed compilation issue * Adjust frequency of trips (#898) * Make network coordinator accessible before injection * Add scenario load hook * Try different version * Frequency adjustment enabled in network coordinator * Trigger new jitpack build * renamed interface and default impl * Fix implementation * Fixing the build using fresh cache * removing costIncludingSubsidy (replacing with cost) (#901) * removing costIncludingSubsidy (replacing with cost) * initial work on getting people to wait for RH manager to do batch allocation next step is deal with how PersonAgent get's a trigger back to go into choosing mode * Fix pt scaling (#903) * Fix pt scaling * Change type of scaling in config to optional * Remove maven.geotoolkit as it breaks build occasionally * person goes to WaitingToChooseMode and waits for a trigger which must be scheduled by RHM * Pass trave time across network (#784) * Added `TravelTimeCalculatorHelper` * Handle `TryToSerialize` and `UpdateTravelTime_v2` in `BeamRouter` * - Handle `UpdateTravelTime_v2` in `R5RoutingWorker` - Fixed tests because need to pass scenario to `BeamRouter` and `R5RoutingWorker` * Log serialized size only if BeamRouter logger is in debug mode * - Renamed `UpdateTravelTime` to `UpdateTravelTimeLocal` - Introduced `UpdateTravelTimeRemote` * Implemented `TravelTimeCalculatorHelper.GetLinkIdToTravelTimeArray` to get map which contains per hour travel time per link * Fixed `maxHour` for `BeamWarmStart` * Removed `return` * - Initialise `travelTimeCalculatorConfigGroup` in R5RoutingWorker - Use `matsimConfig` in `BeamHelper` to get `maxHour` * Removed `TravelTimeDataWithoutLink` * Added `linkIdMap` as immutable collection * Convert `linkIdMap` to `HashMap[Int, Link]` * Fixed runtime error * Use Kryo `4.0.2` * Fixed `WarmStartRoutingSpec` after merge * Dummy change to trigger build * StuckFinder: Better stuck detection (#668) * Detect overall simulation stuck * Add `overallSimulationTimeoutMs` config * Stats about max number of trigger messages per actor type * Fix compilation error for `StuckFinderSpec` * Statistics per trigger type * A bit better way to update `actorToNumOfTriggerMessages` * - generated config - applied `scalafmtAll` * - applied `scalafmtAll` * Added two indexes for fast filter and find * - Extract things related to indexing to IndexerForZonalParkingManager - Use `view` instead of `collectFirst` * Test building indexes in `IndexerForZonalParkingManager` * Increased the level of logs * Added `jfrArgs` * touch to force build * refresh config changes * - Added some gradle functions - Reduced `MetaspaceSize` to 150 Mbytes * Added new trigger type and removed old to/form `triggerTypeThresholds` * Added `beam.debug.stuckAgentDetection.thresholds.actorTypeToMaxNumberOfMessages` * - Fixed `actorTypeToMaxNumberOfMessages` config - Fixed `TriggerMeasurer.getStat` to give correct number of messages - `StuckFinder` now can handle too many trigger message to actor * temp fix for reading from `toll.dat` * Adjust `actorTypeToMaxNumberOfMessages` with data from sf-bay prod run * Added `checkMaxNumberOfMessagesEnabled` to control is check for max number of messages used * Dummy change to trigger build * Rolled back TollCalculator.scala * Set number of agents back to 25000 * Applied `generateConfig` * Add a perf tip when it comes to immutable vs mutable (#915) * Add a perf tip when it comes to immutable vs mutable * Add additional possibility * Trigger new jitpack build * Fix beam-utils dependency to trigger new jitpack build * set tagID instead of master-SNAPSHOT (#918) * set tagID instead of master-SNAPSHOT * bad bad.... mix of new code getting folks to start considering pooled ride hail in choice plus fmt * been a while since we did a fmt * get pooled or not into the request to RHM * refactoring and prep for batch processing requests * dont do that * refactor travel proposal * make it clear you need to turn buffering off with a 0 for interval * refactoring buffered RH requests * Bl/#837 output data documentation 4ci (#923) * new class to describe the fields in any output data file generated * new interface to define a stub to get output data descriptions for each class tht generates an output file * output data descriptors implemented for ModeChosenAnalysis * Merge branch 'master' into bl/#837-output-data-documentation-4ci # Conflicts: # src/main/java/beam/analysis/plots/ModeChosenAnalysis.java * output data descriptors implemented for ModeChosenAnalysis * new class to generate the data descriptions table and write it to an output file * data descriptions added for RealizedModeChoiceAnalysis and RideHailRevenueAnalysis * data descriptions added for PersonTravelTimeAnalysis * data descriptions added for FuelUsageAnalysis * data descriptions added for ExpectedMaxUtilityHeatMap * data descriptions added for physsim stats * data descriptions added for RideHailWaitingAnalysis * data descriptions added for RideHailWaitingAnalysis and Graph Surge Pricing * partial commit * data descriptions added for DeadHeadingAnalysis , summary stats , score stats and stop watch * data descriptions added for summarystats * data descriptions added for events * data descriptions added for rideHailTripDistanceOutputs, tripDurationOutputs, biasErrorGraphDataOutputs, biasNormalizedErrorGraphDataOutputs * Merge branch 'master' into bl/#837-output-data-documentation-4ci # Conflicts: # src/main/scala/beam/sim/config/BeamConfig.scala * use injected transport network while creating BeamMobsim instance in Output data description generator class * been a while since we did a fmt (#922) * been a while since we did a fmt * fmt * getting to the point where we can actually dispatch vehicles to multi-passengers * fmt * configuring API for batch request processing including how to handle less structured route requests * make it clear that matsimConversion is from repo not releases * add config param beam output writeGraphs (#925) * add parameter to config beam.outputs.writeGraphs * add parameter to config beam.outputs.writeGraphs * revert number of last iterations * code is updated after removing param in beam.conf * hold createGraphs value in variable for further use * Re-activate lots of tests; simplify; analyse tolls (#929) * Re-enable a test * Re-enable another test * Re-enable another test * Re-enable another test * Re-enable another test * Remove duplicate functionality * Simplify * Fix test again * Work on OtherPersonAgentSpec * Work on OtherPersonAgentSpec * Work on SingleModeSpec * Remove PWD variable in beam.conf * Reactivate OtherPersonAgentSpec * Reactivate SingleModeSpec * People can "walk" even in the "walk_transit" case * RideHailIterationHistory shouldn't be an Actor * Work on tests * Add missing GTFS file * Report paid amounts in PathTraversalEvent * Re-enable test * Limit Mockito's memory usage * Remove unused class parameters * Put a fare in PersonAgentSpec * Pass toll from driving behavior to "parent" agent * Ensure tolls reach personagent and cost events are thrown * TollRevenueAnalysis * Reformat * Reformat * Merge * Update caches * Spelling; cosmetics * make dummy.zip available from source that isnt gitlfs * starting to unify interface for allocation between single vs batched * more refactoring to unify APIs * more work to process routingRequiredForAllocation but still having trouble with batch versus single processing * Don't stop transit drivers prematurely (#941) * Re-arrange messages in test for (hopefully) deterministic outcome * Test a transit-using Person in ensemble with two TransitDrivers * "Expect" the events from the agents separately (Events from different agents could interleave..) * Don't stop transit drivers prematurely * Format * Put the BeamVille toll on a link most people use * Remove unused code * add log warning for benchmarkFileLoc (#942) * major refactor of ride hail protocols, still needs some TODOs addressed and debugging * things seem to function now up until createTriggersFromMatchedVehicles and we still are not using cached TravelProposals * Remove final scala warnings and enable warning as error (#914) * Remove final scala warnings and enable warning as error * another warning as error * Fix new warning * structuring alloc response to allow pooling * Cs/#858 parking cost in mode choice (#937) * about 3/4 of the way to functioning integration of parking in mode choice * complete steps to get data into chooses mode process * catch unhandled parking inquiry response * conflict * debugging and finalizing consideration of parking cost in mode choice * #858 - update expensive parking costs, parkingSpec passing with reduced iterations * fmt * conflict * test still breaks periodically * now we are no longer missing splitLegForParking on drive transit where drive is the egress leg and we also now are correctly adding tolls when driving on egress, was just 0 before * more refactoring to facilitate manaing routing request / ride hail request mappings * prevent throwing person cost event when the cost is equal to 0.0 (#950) * prevent throwing personCost event when the cost is equal to 0.0 * fix failing tests * fmt and comments * move debugging code out of the way * minor refactoring and simplification of code * simplifying and refactoring for clarity * put modify pass sched manager protocol into docs * final tweak to diagram * better show persons role * simplify and prepare for final stage of reserving rides * add fare adjustments in r5 (#949) * updated subsidies for agency and route * updated map name * updated filter to filterKeys * remove duplications and added comments to code * test added * test added * added null check for tests * updated tests * updated tests * test updated * more cleanup * get everything to run at least through RideHailResponse * simplifying beamville for development * fmt * fmt * touched debug file * import * getting back to a place where runs can happen though still not working * R5RoutingWorker: Override travel time (#948) * Added `TravelTimeAndCost` trait * Use `TravelTimeAndCost` in `BeamRouter` and `R5RoutingWorker` * Implemented case when we have travel time * Implemented case when we have cost * Fixes after solving merge conficts * Zb/#892 add fare adjustments in r5 revert 4ci (#955) * Revert "R5RoutingWorker: Override travel time (#948)" This reverts commit 53073e3. * Revert "add fare adjustments in r5 (#949)" This reverts commit db97fe6. * reverted fare adjustment * debugging bringing functionality back * get rid of BeamTime once and for all, sheesh * tweaks * getting batched allocation to run once through, now need to get unbatched working again * make sure all completes happen at correct time * debugging * return transti * debug * catch case when routes are empty * Harmonize event field names (#956) * Harmonize event field names * fix test * Bl/#848 remove unused config options 4ci (#924) * remove unused keys in config files - beam.basePackage - beam.agentsim.agents.drivingCost.defaultLitersPerMeter - beam.agentsim.agents.drivingCost.defaultPricePerGallon - beam.agentsim.agents.rideHail.vehicleRangeInMeters - beam.debug.skipOverBadActors = true - beam.debug.secondsToWaitForSkip = 30 * remove beam.debug.skipOverBadActors usage in RideHailTestHelper * removed akka.conf from sfbay prod.conf * remove production folder entirely * regenerated beam config * fixed compilation issues with removed config options * make default allocation behavior work * bring params closer to always passing status * fixing bugs getting tests to pass * remove accessMode from beamTrip (#967) * pt fares (#962) * pt fares * pt fares, test fixes * updated tests * updated vehicle id * Add new scalaFmt check for CI and run scalaFmtAll (#971) * Add new scalaFmt check for CI and run scalaFmtAll * configuration consistency check (#945) * write code for configuration consistency check * resolved compilation issue on jenkins * log formatting and remove datatype * remove unnecessary variable * remove key values string before property and values respectively * bring params closer to always passing status * improving clarity of output and starting work on making configs more consistent * more work improving message and making consistent * make a default value for clusterType * Bl/#965 graph comparisons issue 4ci (#975) * comparison.html to use relative paths for image src rather than absolute path * wrap the overflowing text in the list items for graph comparison html page * throw exception, when pt fares file set, but does not exist (#976) * #974 throw exception, when pt fares file set, but does not exist * #974 throw exception, when pt fares file set, but does not exist * added ptFares.csv to beamville (#977) * need average car travel time over iterations graph (#970) * create a root iteration graph for average car travel times * change the variable name * change the string literal with variable * update the code for average travel time and fix failing test * get average on the basis of count instead of hours * bug fix: pt fare age rule (#981) * bug fix: pt fare age rule * added subsidy and ptfare for sflight * updated fare amount * #985 throw exception, when subsidies file set, but does not exist (#986) * #983 enable beam to publish build to local maven (#984) * make simulation robust by handling person travel time computation (#982) * #987 beam jar name should always be beam (#989) * #988 PersonCostEvent should raise based on subsidized cost (#990) * debugging * fmt * resolve * Cs/#993 badtest (#994) * change structure of test * bad old stuff * fmt * must be an average * much progress working out how to request routes for pooled rides * dev config * more debugging, now RHM attempts pooling and it runs to completion, but half the agents aren't entering vehicle * they pool * fixing mistake in creation of passenger schedule * bug fix and fail fast for time zone issue (#957) * make dummy walk padding robust * fmt plus detailed error message * refactor to make easier to use elsewhere * fmt * add check to R5RoutingWorker constructor * fmt * get person agent spec working again * this test doesn't make sense * fmt * compile errors fixed * scala fmt alias task * get beam.conf back to default * revert mode choice to get this passing * revert all the way back * Cs/add removal strategy (#1000) * add config option so we can test it during fail fast * add first config option check in fail fast (#998) * create a class fail fast and added config option check for fail fast * scala formatting * make fail fast actually fail and add pooling * one more fail fast condition * blah, thought I changed this back * fmt * consistentcy for sf-light configs * deploy example with branch * just 1 iter * update timeout stuck detection classnames * getting it to run in more contexts * update base remove deprecated * more tweaks to get running again * update tolls * more peeps * final cleanup * fmt * sort the list of defaults
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This change is