The wpimath library is a new library designed to separate the reusable math functionality
from the common utility library (wpiutil) and the hardware-dependent library (wpilibc/j).
Package names / include file names were NOT changed to minimize breakage. In a future year
it would be good to revamp these for a more uniform user experience and to reduce the risk
of accidental naming conflicts.
While theoretically all of this functionality could be placed into wpiutil, several pieces
of this library (e.g. DARE) are very time-consuming to compile, so it's nice to avoid this
expense for users who only want cscore or ntcore. It also allows for easy future separation
of build tasks vs number of workers on memory-constrained machines.
This moves the following functionality from wpiutil into wpimath:
- Eigen
- ejml
- Drake
- DARE
- wpiutil.math package (Matrix etc)
- units
And the following functionality from wpilibc/j into wpimath:
- Geometry
- Kinematics
- Spline
- Trajectory
- LinearFilter
- MedianFilter
- Feed-forward controllers
Remove WaitForCachedData as it's no longer required.
Also properly handle caching / transition detection logic that occurs at the
WPILib level.
This also changes DriverStation::IsNewControlData() to check for WPILib-level
caching instead of wrapping the HAL function.
Using GetAverageVoltage will reduce the noise in potentiometer reading. Potentiometers are unlikely to be used where the minimal lag averaging introduces would impact control loop stability.
When not direct mapped, make index constructors private and add factory
functions for channel and index.
Co-authored-by: GabrielDeml <gabrielddeml@gmail.com>
pose.Translation().X() and pose.Translation.Y() are common operations,
so shortening them to pose.X() and pose.Y() would be convenient.
Java uses the getX() convention so that is used instead of X() for Java.
Also move some things in HAL for consistency.
WAS:
C++:
- C APIs: #include "mockdata/AccelerometerData.h"
- User side class: #include "simulation/AccelerometerSim.h"
Java:
- JNI APIs: hal.sim.mockdata.AccelerometerData (and a few classes in hal.sim)
- User side classes: hal.sim.AccelerometerSim
IS:
C++:
- C APIs: #include "hal/simulation/AccelerometerData.h"
- C++ class: #include "frc/simulation/AccelerometerSim.h"
Java:
- JNI APIs: hal.simulation.AccelerometerData
- User side class: wpilibj.simulation.AccelerometerSim
The most common mistake users (including contributors to WPILib) seem to make while creating new constraints is ignoring some sort of edge case that causes the calculated minimum acceleration to be greater than the calculated maximum acceleration.
This specialized exception, with its detailed error message, should make it easier and quicker for said users to debug and fix bugs within their constraints.
Co-authored-by: Tyler Veness <calcmogul@gmail.com>
If the 64 bit FPGA timer rolls over, a 32 bit value is added for
the rollover, an artifact of when it was a 32 bit timer.
The 64 bit microsecond timer won't rollover for 500k years so remove the
check for simplicity.
Fixes#2504
Valgrind caught this one. If you default initialize a Trajectory, it
doesn't initialize m_totalTime. This means a subsequent call to
Trajectory::TotalTime() returns the value of an uninitialized variable.
3512's software was using 0_s as a sentinel value for when a trajectory
was empty. We didn't notice a problem before because Linux zero-inits
memory.
Using this overload makes the thread backing the Notifier run at
real-time priority. This improves scheduling jitter substantially (5ms
+- 2ms down to 5ms +- 1ms).
A version isn't provided for Java because making threads real-time can
cause GC deadlocks.
The DutyCycleEncoder class initializes AnalogTrigger, which is not supported in simulation.
To avoid this, do not use AnalogTrigger (or Counter) in simulation mode.
Fixes#2367
Co-authored-by: Peter Johnson <johnson.peter@gmail.com>
This was caused by m_epochs storing the timestamps as nanoseconds while
the epoch printing code expects microseconds. Adding a duration cast
fixes this.
Java stores the epoch timestamps in a double as microseconds, so it
doesn't exhibit this bug. The comments were updated to make this more
obvious.
Fixes#2392.
Previously, it could take the long way around. This recomputes the
profile goal with the shortest error, thus taking the shortest path.
Also removed the setpoint clamping from PIDController::SetSetpoint()
because it's unnecessary to make PIDController behave correctly for
a modular arithmetic input, and it breaks the setpoint calculation in
ProfiledPIDController otherwise.
Fixes#2277.
