The full Rust API is on docs.rs. This page says where to start. To get everything through one
dependency, use dspk, which re-exports the crates below as modules.
| Crate |
What it is |
dspk |
Entry crate: re-exports the others; features choose what’s included |
dspk-core |
Recordings, buffers, exact time and sample rates, windows, compute targets |
dspk-io |
Raw binary and Zarr recordings, format detection, chunk cache |
dspk-ops |
Filters, resampling, spatial operators, linear algebra, statistics, peaks, device pipelines |
dspk-pipeline |
Pipelines as data: plan, many inputs, merges, provenance |
dspk-view |
Min/max envelopes and pyramids for viewing |
dspk-stream |
Network sessions for continuous signals (QUIC + TLS, protobuf) |
dspk-hub |
Verified downloads of published artifacts |
dspk-synapse |
Electrophysiology: formats, probes, detection, clustering, matching, metrics |
dspk-synapse-ml |
Pretrained models and their runtime |
dspk-sorter-kilosort |
Kilosort4 and EMUsort (GPL-3.0) |
dspk-sorter-ms5 |
MountainSort 5 (Apache-2.0) |
dspk-sorter-si |
SpyKING CIRCUS 2 and Tridesclous 2 |
dspk-app-core |
The app’s core and its plugin interface |
| Need |
Module |
Start with |
| Chain preprocessing steps |
pipeline |
Pipeline, PipelineStage |
| Filters |
filter |
FilterSpec, execute_filter |
| Mix channels |
spatial |
SpatialWhitening, SurfaceLaplacian |
| Noise, percentiles, element-wise math |
math |
estimate_noise_std, execute_channel_mad |
| Matrix products, PCA, ICA |
linalg |
PcaModel, matmul |
| Peaks |
peaks |
find_peak_candidates |
| Resampling |
resampler |
execute_resample_poly |
- Data is a
[channels, samples] buffer, channel-major.
- Functions take a CubeCL
Client: ComputeTarget::from_env()?.client()? picks the runtime named by
DSPK_RUNTIME, or else the first compiled in.
- Functions on device buffers start with
execute_. Data comes first with its shape, then
sample_rate, then one settings struct.
