Add type hints and dataclasses for metadata

pyproject.toml

@@ -5,29 +5,21 @@ build-backend = "setuptools.build_meta"
 [project]
 name = "open-ephys-python-tools"
 description = "Software tools for interfacing with the Open Ephys GUI"
-license = {text = "MIT"}
-requires-python = ">=3.7"
+license = { text = "MIT" }
+requires-python = ">=3.9"
 classifiers = [
     "Programming Language :: Python :: 3",
     "License :: OSI Approved :: MIT License",
-    "Operating System :: OS Independent"
+    "Operating System :: OS Independent",
 ]
 readme = "README.md"
 
 dynamic = ["version"]
 
-dependencies = [
-    'numpy',
-    'pandas',
-    'h5py',
-    'zmq',
-    'requests'
-]
+dependencies = ['numpy', 'pandas', 'h5py', 'zmq', 'requests']
 
 [tool.setuptools.packages.find]
 where = ["src"]
 
 [tool.setuptools.dynamic]
-version = {attr = "open_ephys.__version__"}
-
-
+version = { attr = "open_ephys.__version__" }

src/open_ephys/analysis/README.md

@@ -72,7 +72,7 @@ Each `continuous` object has four properties:
 - `samples` - a `numpy.ndarray` that holds the actual continuous data with dimensions of samples x channels. For Binary, NWB, and Kwik format, this will be a memory-mapped array (i.e., the data will only be loaded into memory when specific samples are accessed).
 - `sample_numbers` - a `numpy.ndarray` that holds the sample numbers since the start of acquisition. This will have the same size as the first dimension of the `samples` array
 - `timestamps` - a `numpy.ndarray` that holds global timestamps (in seconds) for each sample, assuming all data streams were synchronized in this recording. This will have the same size as the first dimension of the `samples` array
-- `metadata` - a `dict` containing information about this data, such as the ID of the processor it originated from.
+- `metadata` - a `ContinousMetadata` dataclass containing information about this data, such as the ID of the processor it originated from.
 
 Because the memory-mapped samples are stored as 16-bit integers in arbitrary units, all analysis should be done on a scaled version of these samples. To load the samples scaled to microvolts, use the `get_samples()` method:
 
@@ -124,7 +124,7 @@ If spike data has been saved by your Record Node (i.e., there is a Spike Detecto
 - `sample_numbers` - `numpy.ndarray` of sample indices (one per spikes)
 - `timestamps` - `numpy.ndarray` of global timestamps (in seconds)
 - `clusters` - `numpy.ndarray` of cluster IDs for each spike (default cluster = 0)
-- `metadata` - `dict` with metadata about each electrode
+- `metadata` - `SpikeMetadata` dataclass with metadata about each electrode
 
 ## Synchronizing timestamps