v0.18.0

TestParticle v0.18.0

Diff since v0.17.0

Breaking

• Species definition has been refactored; User type removed
• Boris methods now return ODESolution instead of previously natively defined TraceSolution
• Removed native VDF definitions; now switched to VelocityDistributionFunctions.jl

Features

• Added Unitful support
• Added PartitionArray support for Unitful
• Added get_number_density and get_number_density_flux utilities

Merged pull requests:

• Update shock demo to use Axis3 and km units (#363) (@henry2004y)
• Refactor CurrentLoop implementation for arbitrary orientation (#366) (@henry2004y)
• Refactor trace_fieldline to accept various field types directly (#367) (@henry2004y)
• Refactor distributions (#368) (@henry2004y)
• Add demo for proton motion in Earth gravity (#370) (@henry2004y)
• VelocityDistributionFunctions related update (#372) (@Beforerr)
• Unitful support (#373) (@Beforerr)
• Update sampling demo to use new VDF interface (#374) (@henry2004y)
• feat!: remove the need to specify User when q and m are specified. (#375) (@Beforerr)
• refactor: cleanup unnessecary annotation (#376) (@Beforerr)
• feat: Add ArrayPartition support for heterogeneous arrays (#378) (@Beforerr)
• Add flux density calculation and sphere sampling utilities
  (#381) (@henry2004y)
• Add get_number_density utility function (#382) (@henry2004y)
• Update shock demo to use new sampling API (#383) (@henry2004y)
• docs: fix benchmark (#384) (@Beforerr)
• build(deps): bump actions/cache from 4 to 5 (#385) (@dependabot[bot])
• build(deps): bump peter-evans/create-pull-request from 7 to 8 (#386) (@dependabot[bot])
• Add AutoVern series to performance benchmark demo (#388) (@henry2004y)
• Convert demo_interpolation.jl to markdown and enable benchmarks (#389) (@henry2004y)
• Update performance demo to use log scale and mention defaults (#390) (@henry2004y)
• Add demo for efficient storage of ensemble solutions (#391) (@henry2004y)
• Refactor Boris solvers to use SciMLBase.ODESolution and optimize with @muladd (#392) (@henry2004y)

Closed issues:

• Add more examples (#19)
• Consolidate efforts about analytic velocity distribution (#364)
• Displaying multiple benchmark results (#379)

v0.17.0

TestParticle v0.17.0

Diff since v0.16.2

Breaking changes

• Refactor grid interpolation to use Meshes.jl types; no more internal grid types

Merged pull requests:

• Optimize plot_dist function in demo_fermi_foreshock.jl (#360) (@henry2004y)
• Refactor grid interpolation to use Meshes.jl types (#361) (@henry2004y)
• Resolve test discrepancy in trace function comparison (#362) (@henry2004y)

v0.16.2

TestParticle v0.16.2

Diff since v0.16.1

• Introduce DipoleField and HarrisCurrentSheet type setup
• Substitute isoutofdomain with DiscreteCallback
• Fix compatibility with Meshes 0.56

Merged pull requests:

• Update FLR demo and refactor GC drift equations (#350) (@henry2004y)
• Add Z-pinch particle tracing demo (#351) (@henry2004y)
• doc: use callback instead of isoutofdomain (#352) (@Beforerr)
• Better magnetic field configuration (#353) (@Beforerr)
• Replace isoutofdomain usage with DiscreteCallback (#354) (@henry2004y)
• Update Magnetic Dipole demo to use Axis3 and Re units (#356) (@henry2004y)
• Update Analytical Magnetosphere demo visualization to use Axis3 (#357) (@henry2004y)
• Add Shock Drift Acceleration demo (#358) (@henry2004y)
• Update CartesianGrid constructor for Meshes.jl v0.56.0 (#359) (@henry2004y)

Closed issues:

• Improve the FLR example (#59)
• Refactor and add magnetostatics features (#165)
• Add pinch configurations (#192)

v0.16.1

TestParticle v0.16.1

Diff since v0.16.0

• Optimize memory allocations in Boris and Multistep Boris solvers
• Add adaptive Boris method

Merged pull requests:

• Fix flaky BiKappa test by increasing sample size (#344) (@henry2004y)
• Fix incorrect time extraction in get_gc for 6-element state vectors (#345) (@henry2004y)
• Optimize memory allocations in Boris and Multistep Boris solvers (#348) (@henry2004y)
• feat: Adaptive Boris method (#349) (@henry2004y)

Closed issues:

• Adaptive timestepping control (#131)
• Changing dt for the Boris method (#144)
• Time in guiding center (#213)

v0.16.0

TestParticle v0.16.0

Diff since v0.15.4

• Refactor sampling and add support for kappa and self-similar distribution
• Reorganized documentation demonstrations
• Improved Boris pusher with refined control

Breaking changes

• Distribution types now save physical thermal speed with a factor 2.

Merged pull requests:

• Refactor Boris performance demo into a standalone example (#325) (@henry2004y)
• Improve performance benchmark realism and visualization (#326) (@henry2004y)
• Add loss cone distribution visualization to magnetic mirror demo (#327) (@henry2004y)
• Add demo for adiabatic invariants periods (#328) (@henry2004y)
• feat: Add getB_loop for current coil magnetic fields (#329) (@henry2004y)
• Refactor energy ratio comparison in dipole demo to use a table (#331) (@henry2004y)
• Update performance demo with Markdown table and reduced plot gap (#333) (@henry2004y)
• Add Kappa and Self-similar distribution functions (#334) (@henry2004y)
• Add magnetic field line tracing support and update gitignore (#335) (@henry2004y)
• Add Reconnection demo (#336) (@henry2004y)
• Increase particle tracing times in x-line reconnection demo (#338) (@henry2004y)
• doc: add field tracing demo (#340) (@henry2004y)
• Optimize native pushers for time-independent fields (#341) (@henry2004y)
• Refactor Maxwellian and BiMaxwellian sampling to use randn (#342) (@henry2004y)

Closed issues:

• Add an adiabatic invariants demo (#156)
• Field line tracing along trajectories (#162)
• Add a demo for reconnection (#163)
• Pitch angle distribution (#190)
• Add more distribution setup (#191)
• Boris method optimization (#308)

v0.15.4

TestParticle v0.15.4

Diff since v0.15.3

• Fixed the Boris solver output velocity mismatch.
• Fixed the NaN boundary interpolation bug.

Merged pull requests:

• Sync velocity output in Boris method (#321) (@henry2004y)
• Refactor demo_flux.jl to separate examples and improve structure (#322) (@henry2004y)
• fix: NaN boundary dimension mismatch (#323) (@henry2004y)

Closed issues:

• Initial step energy error (#309)
• Interpolation NaN boundary dimension mismatch (#313)

v0.15.3

TestParticle v0.15.3

Diff since v0.15.2

• Optimized vector field interpolation
• Support nonuniform Cartesian grid interpolation
• Add multi-step Boris pusher
• Fine-tuning of the Boris methods

Merged pull requests:

• Update demo_energy_conservation.jl with new test cases and solvers (#298) (@henry2004y)
• Fix energy conservation demo setup, stability, and comments (#299) (@henry2004y)
• Fix the phi range from 0 to 2pi (#302) (@henry2004y)
• Improve phi direction interpolation for SphericalNonUniformR and use in-place interpolate! (#303) (@henry2004y)
• Optimize vector field interpolation using SVector and reinterpret (#304) (@henry2004y)
• Modify the Boris 2-thread test (#305) (@henry2004y)
• Fix Boris pusher initialization using explicit Euler backward step (#306) (@henry2004y)
• Synchronize Boris pusher output velocity (#307) (@henry2004y)
• Reduce allocations in spherical uniform interpolation (#311) (@henry2004y)
• Switch Boris initialization back to backward Boris step (#312) (@henry2004y)
• Fix final velocity saving in Boris pushers with explicit step calculation (#315) (@henry2004y)
• Update demos with marker plots and correct phase error order
  (#316) (@henry2004y)
• Update energy conservation demo with variable dt and refactor (#317) (@henry2004y)
• Add save_everystep and save_end keywords to solve (#318) (@henry2004y)
• Add CartesianNonUniform grid type for interpolation (#319) (@henry2004y)

Closed issues:

• Energy conservation test (#73)
• Multithreading Boris method race condition? (#130)
• Save the final timepoint for Boris method (#194)
• Generalize interpolation for irregular grid (#214)
• Initial step energy error (#309)

v0.15.2

TestParticle v0.15.2

Diff since v0.15.1

Merged pull requests:

• Upgrade documentation to use DocumenterVitepress (#276) (@henry2004y)
• Move documentation generation to a separate action (#277) (@henry2004y)
• Debug document generation (#278) (@henry2004y)
• Doc demos cleanup (#281) (@henry2004y)
• Fix dead cross-references in documentation demos (#282) (@henry2004y)
• doc: small adjustments (#283) (@henry2004y)
• Reorganize documentation examples into semantic categories (#284) (@henry2004y)
• Docs: Add demo for InterpolatedDiskArray (#285) (@henry2004y)
• Consolidate dimensionless unit demos and unify title capitalization (#287) (@henry2004y)
• Combine Boris solver demos into a single documentation example (#288) (@henry2004y)
• Improve particle flux demo consistency and conciseness (#289) (@henry2004y)
• Merge ensemble tracing demos (#290) (@henry2004y)
• Update docstrings for gyro-related utility functions (#291) (@henry2004y)
• feat: multistep Boris solver (#292) (@henry2004y)
• Bump actions/checkout from 4 to 6 (#293) (@dependabot[bot])
• doc: Multistep boris demo (#294) (@henry2004y)
• Add phase error analysis demo (#296) (@henry2004y)

Closed issues:

• Phase error tests (#132)
• Multistep Boris Solver (#253)

v0.15.1

TestParticle v0.15.1

Diff since v0.15.0

Merged pull requests:

• Refactor interpolations with better support for spherical mesh (#271) (@henry2004y)
• doc: add radiation demo (#273) (@henry2004y)
• Merge imports (#274) (@henry2004y)

Closed issues:

• Unstable tests (#270)

v0.15.0

TestParticle v0.15.0

Diff since v0.14.2

Breaking changes

• Introduced grid geometry for the field interpolation. Cartesian() is set for backward compatibility. Spherical() and SphericalNonUniformR() are introduced.
• The transformation between the Cartesian coordinates and the spherical coordinates now follows the (r, theta, phi) convention.

Merged pull requests:

• Bump actions/checkout from 5 to 6 (#262) (@dependabot[bot])
• feat: Add non-uniform spherical interpolation and optimizations (#266) (@henry2004y)
• Fix spherical interpolation theta BC (#268) (@henry2004y)