Skip to content

Releases: jaxleyverse/jaxley

v0.11.3

23 Aug 08:52
e76476f
Compare
Choose a tag to compare

🛠️ Internal updates

🐛 Bug fixes

  • fix for networks that mix point neurons and morphologically detailed neurons (#702,
    @michaeldeistler)
  • carry over groups from constituents of a module (e.g., jx.Cell groups get carried
    over to jx.Network) (#703, @michaeldeistler)

v0.11.2

14 Jul 12:47
2c2a3da
Compare
Choose a tag to compare

🐛 Bug fixes

  • Bugfix for Networks on GPU: since v0.9.0, networks had been very slow on GPU
    because the voltage equations of cells had been processed in sequence, not in parallel.
    This is now solved, giving a large speed-up for networks consisting of many cells (#691,
    @michaeldeistler, thanks to @VENOM314 for reporting)

📚 Documentation

v0.11.1

10 Jul 16:10
b5d0436
Compare
Choose a tag to compare

🐛 Bug fixes

🛠️ Internal updates

v0.11.0

04 Jul 16:41
d33042f
Compare
Choose a tag to compare

🧩 New features

📚 Documentation

🛠️ Internal updates

  • changes to how the membrane area from SWC files is computed when the radius within a
    compartment is not constant. This fix can have an impact on simulation results. The
    updated computation of membrane area matches that of the NEURON simulator (#662,
    @michaeldeistler, thanks to @VENOM314 for reporting).
  • remove the custom SWC reader (which had been deprecated in v0.9.0, #662,
    @michaeldeistler).
  • fix bug in set_ncomp (# 676, @manuelgloeckler)

v0.10.0

11 Jun 07:44
24dd530
Compare
Choose a tag to compare

🧩 New features

from jaxley.morphology import distance_pathwise
path_dists = distance_pathwise(cell.soma.branch(0).comp(0), cell)
cell.nodes["path_dist_from_soma"] = path_dists

🐛 Bug fixes

🛠️ Internal updates

  • module.distance() is now deprecated in favor of jx.morphology.distance_direct()
    (#648, @michaeldeistler)

v0.9.0

23 May 10:37
6af8040
Compare
Choose a tag to compare

✨ Highlights

  • This PR implements a new solver, which is now used by default (#625,
    @michaeldeistler). The new solver has the following advantages:
    • Much lower runtime. Across several morphologies, we get a 20% runtime speedup on
      CPU and a 50% runtime speedup on GPU.
    • Almost zero compile time. We get a 50x compile time speedup on CPU and a 3x compile
      time speedup on GPU.
  • Utility to delete parts of a morphology (#612, @michaeldeistler):
from jaxley.morphology import morph_delete
cell = morph_delete(cell.branch([1, 2]))
from jaxley.morphology import morph_connect
cell = morph_connect(cell1.branch(1).loc(0.0), cell2.branch(2).loc(1.0))

🧩 New features

  • the default SWC reader has changed. To use the previous SWC reader, run
    jx.read_swc(..., backend="custom"). However, note that we will remove this reader
    in the future. If the new SWC reader is causing issues for you, please open an issue
    (#612, @michaeldeistler)
  • radiuses are now integrated across SWC coordinates, not interpolated
    (#612, @michaeldeistler)
  • remove pin of JAX version. New JAX versions (JAX>=0.6.0) resolve slow CPU
    runtime, see here (#623, @michaeldeistler)
  • running the d_lambda rule is now much faster, see
    the how-to guide
    (#625, @michaeldeistler)

📚 Documentation

🛠️ Internal updates

🐛 Bug fixes

🎉 New Contributors

v0.8.2

27 Mar 15:02
f9fc4f8
Compare
Choose a tag to compare

v0.8.1

13 Mar 17:23
143b9a2
Compare
Choose a tag to compare

v0.8.0

13 Mar 16:03
52c11a7
Compare
Choose a tag to compare

New features

Minor updates

  • make delta and v_th in IonotropicSynapse trainable parameters (#599, @jnsbck)
  • make random postsnaptic compartment selection optional in connectivity functions (#489, @kyralianaka)

Bug fixes

v0.7.0

26 Feb 13:51
70bf60c
Compare
Choose a tag to compare

New Features

from jaxley.channels import CaNernstReversal
from jaxley.pumps import CaFaradayConcentrationChange

cell.insert(CaFaradayConcentrationChange())
cell.insert(CaNernstReversal())
cell.diffuse("CaCon_i")
cell.set("axial_diffusion_CaCon_i", 1.0)

Bug fixes

Code health