`canvod.auxiliary` API reference¶

SP3 ephemeris and CLK clock correction processing, interpolation, coordinate transformations, and the GNSS product registry.

Package¶

canvod-aux: Auxiliary data augmentation for GNSS VOD analysis

Handles downloading, parsing, and interpolating SP3 ephemerides and clock corrections for GNSS satellite data processing.

`Sp3File` ¶

Bases: AuxFile

Handler for SP3 orbit files with multi-product support.

Now supports all IGS analysis centers via product registry: COD, GFZ, ESA, JPL, IGS, WHU, GRG, SHA

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True.

Attributes:

Name	Type	Description
`date`	`str`	String in YYYYDOY format.
`agency`	`str`	Agency code (e.g., "COD", "GFZ", "ESA").
`product_type`	`str`	Product type ("final", "rapid").
`ftp_server`	`str`	Base URL for downloads.
`local_dir`	`Path`	Local storage directory.
`add_velocities`	`bool \| None, default True`	Whether to compute velocities.
`dimensionless`	`bool \| None, default True`	Whether to strip units (store magnitudes only).
`product_spec`	`(ProductSpec \| None, optional)`	Product specification resolved from the registry.

`__post_init__()` ¶

Initialize with product validation.

`get_interpolation_strategy()` ¶

Get appropriate interpolation strategy for SP3 files.

`candidate_filenames()` ¶

SP3 file names, one per sampling the product has been published with.

Pattern: {PREFIX}{YYYYDOY}0000_01D_ORB.SP3

Example: COD0MGXFIN_20240150000_01D_05M_ORB.SP3

`download_aux_file(filename)` ¶

Download one SP3 file, trying all servers from the product registry.

Servers are tried in priority order from products.toml. Auth-required servers are skipped when no credentials are configured. Warnings are printed when falling back to alternate servers.

Parameters:

Name	Type	Description	Default
`filename`	`str`	One of :meth:`candidate_filenames`.	required

Raises:

Type	Description
`RuntimeError`	If download fails from all available servers.
`ValueError`	If GPS week calculation fails.

`read_file()` ¶

Read and validate SP3 file.

Returns:

Type	Description
`Dataset`	Dataset with satellite positions (X, Y, Z) in meters.

Raises:

Type	Description
`FileNotFoundError`	If file does not exist.
`ValueError`	If validation fails.

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

Like :meth:AuxFile.join_days, with the velocities recomputed.

The velocities of each file have one-sided differences at its first and last epoch; recomputed on the joined data, the day's first and last epochs get central differences.

`compute_velocity(ds)` ¶

Compute satellite velocities from position data.

Uses central differences for interior points, forward/backward differences for endpoints, with the actual epoch spacing (the days joined by :meth:join_days may come with different samplings).

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset with X, Y, Z coordinates.	required

Returns:

Type	Description
`Dataset`	Dataset augmented (or updated) with Vx, Vy, Vz velocities.

`ClkFile` ¶

Bases: AuxFile

Handler for GNSS clock files in CLK format.

This class reads and processes clock offset files containing satellite clock corrections. It handles the parsing of CLK format files and provides the data in xarray Dataset format.

Supports multiple analysis centers via product registry with proper FTP paths and filename conventions.

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True.

Attributes:

Name	Type	Description
`date`	`str`	String in YYYYDOY format representing the start date.
`agency`	`str`	Analysis center identifier (e.g., "COD", "GFZ").
`product_type`	`str`	Product type ("final" or "rapid").
`ftp_server`	`str`	Base URL for file downloads.
`local_dir`	`Path`	Local storage directory.
`dimensionless`	`bool \| None, default True`	If True, outputs magnitude-only values (no units attached).

`__post_init__()` ¶

Initialize CLK file handler.

`get_interpolation_strategy()` ¶

Get appropriate interpolation strategy for CLK files.

`candidate_filenames()` ¶

CLK file names, one per sampling the product has been published with.

Filename format: {PREFIX}{YYYYDOY}0000_01D_CLK.CLK

Raises:

Type	Description
`UnknownProductError`	If agency/product combination not in registry.
`ValueError`	If the product publishes no clock files.

`download_aux_file(filename)` ¶

Download one CLK file, trying all servers from the product registry.

Servers are tried in priority order from products.toml. Auth-required servers are skipped when no credentials are configured. Warnings are printed when falling back to alternate servers.

Parameters:

Name	Type	Description	Default
`filename`	`str`	One of :meth:`candidate_filenames`.	required

Raises:

Type	Description
`RuntimeError`	If file cannot be downloaded from any available server.
`ValueError`	If GPS week calculation fails.

`read_file()` ¶

Read and parse CLK file into xarray Dataset.

Uses modular parser for data extraction and validator for quality checks. The clock bias is in seconds (RINEX clock 3.04, Table A16).

Returns:

Type	Description
`Dataset`	Clock offsets with dimensions (epoch, sv). Values are in seconds (or dimensionless if specified).

`AuxFile` ¶

Bases: ABC

Abstract base class for GNSS auxiliary files (SP3, CLK, IONEX, etc.).

This class provides two ways to create instances: 1. from_datetime_date(): Create from a datetime.date object and metadata 2. from_file(): Create directly from an existing file path

The class handles both newly downloaded files and existing local files, maintaining consistent behavior regardless of how the instance is created.

FTP Server Configuration:

user_email: Optional email for NASA CDDIS authentication
If None: Uses ESA FTP server exclusively (no authentication required)
If provided: Enables NASA CDDIS as fallback server (requires registration)
To enable CDDIS, set processing.credentials.nasa_earthdata_acc_mail in config/canvod-settings.yaml

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True, and it uses ABC to define required subclass hooks.

`data` `property` ¶

Access the file's data, loading it if necessary.

`__post_init__()` ¶

Initialize after dataclass creation.

Sets up paths, downloader, and verifies local file existence.

`from_datetime_date(date, agency, product_type, ftp_server, local_dir, **kwargs)` `classmethod` ¶

Create an AuxFile instance from a datetime.date.

Parameters:

Name	Type	Description	Default
`date`	`date`	Date for the desired auxiliary file.	required
`agency`	`str`	Agency providing the data (e.g., "COD", "IGS").	required
`product_type`	`str`	Product type ("final" or "rapid").	required
`ftp_server`	`str`	Base URL for file downloads.	required
`local_dir`	`Path`	Directory for storing files locally.	required
`**kwargs`	`Any`	Extra keyword arguments for subclass construction.	`{}`

Returns:

Type	Description
`AuxFile`	A new instance of the AuxFile subclass.

`from_file(fpath, **kwargs)` `classmethod` ¶

Create an AuxFile instance from an existing file path.

Parameters:

Name	Type	Description	Default
`fpath`	`Path`	Path to the existing GNSS file.	required
`**kwargs`	`Any`	Extra keyword arguments for subclass construction.	`{}`

Returns:

Type	Description
`AuxFile`	A new instance of the AuxFile subclass.

Raises:

Type	Description
`FileNotFoundError`	If the specified file does not exist.
`UnknownProductError`	If no registry product has the prefix of the file name.

`download_file(url, destination, file_info=None)` ¶

Download a file using the configured downloader.

Parameters:

Name	Type	Description	Default
`url`	`str`	Download URL.	required
`destination`	`Path`	Local file destination.	required
`file_info`	`dict`	Extra info passed to the downloader.	`None`

Returns:

Type	Description
`Path`	Path to the downloaded file.

`download_with_fallback(ftp_path, destination, file_info=None, product_spec=None)` ¶

Download a file, trying all servers from the product spec in priority order.

Iterates through the product's FTP server list (from products.toml), skipping servers that require authentication when no credentials are configured. Warns when falling back to an alternate server. Raises a clear error when all servers fail.

Parameters:

Name	Type	Description	Default
`ftp_path`	`str`	Server-relative path (e.g. `/gnss/products/2345/FILE.gz`).	required
`destination`	`Path`	Local file destination.	required
`file_info`	`dict`	Extra context passed to the downloader.	`None`
`product_spec`	`ProductSpec`	Product specification with `ftp_servers` list. If None, falls back to single-server download using `self.ftp_server`.	`None`

Returns:

Type	Description
`Path`	Path to the downloaded file.

Raises:

Type	Description
`RuntimeError`	If download fails from all available servers.

`read_file()` `abstractmethod` ¶

Read and parse the auxiliary file.

Returns:

Type	Description
`Dataset`	Parsed dataset representation of the file.

`get_interpolation_strategy()` `abstractmethod` ¶

Get the interpolation strategy for this file type.

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

This file's data with the end of the previous and the start of the next day.

A daily file can end before midnight (e.g. a 15 min orbit file at 23:45), and its first and last epochs have neighbors on one side only. With margin of the neighboring days on both sides, the whole day lies inside the data: nothing is extrapolated.

Parameters:

Name	Type	Description	Default
`previous`	`AuxFile`	The same product for the previous and the next day.	required
`following`	`AuxFile`	The same product for the previous and the next day.	required
`margin`	`timedelta64`	How much of each neighboring day is kept.	`DAY_JOIN_MARGIN`

Returns:

Type	Description
`Dataset`	The epochs of this file, plus those of the neighbors within `margin` of this day that this file lacks (e.g. this file's 24:00 epoch wins over the next file's 00:00), sorted by epoch. Satellites missing in a file are NaN there. Attrs of this file.

`check_file_exists()` ¶

Return the local file, downloading it if needed.

The candidate names (one per sampling the product has been published with) are tried in order: first a local copy of any of them, then a download of each in turn.

Returns:

Type	Description
`Path`	Local file path.

Raises:

Type	Description
`RuntimeError`	If no candidate could be downloaded; the message names every file tried.

`candidate_filenames()` `abstractmethod` ¶

File names of this file, in the order they are tried.

`download_aux_file(filename)` `abstractmethod` ¶

Download one candidate file into local_dir.

Parameters:

Name	Type	Description	Default
`filename`	`str`	One of :meth:`candidate_filenames`.	required

`Interpolator` ¶

Bases: ABC

Abstract base class for interpolation strategies.

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True and uses ABC to define required interpolation hooks.

`interpolate(ds, target_epochs)` `abstractmethod` ¶

Interpolate dataset to match target epochs.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Source dataset with (epoch, sid) dimensions.	required
`target_epochs`	`ndarray`	Target epoch grid (datetime64).	required

Returns:

Type	Description
`Dataset`	Interpolated dataset at target epochs.

`to_attrs()` ¶

Convert interpolator to attrs-compatible dictionary.

`InterpolatorConfig` ¶

Base class for interpolator configuration.

`to_dict()` ¶

Convert config to dictionary for attrs storage.

`DatasetMatcher` ¶

Match auxiliary datasets to a reference RINEX dataset temporally.

Handles temporal alignment of datasets with different sampling rates using appropriate interpolation strategies. The reference dataset (typically RINEX observations) remains unchanged while auxiliary datasets are interpolated to match its epochs.

The matcher: 1. Validates all datasets have required dimensions (epoch, sid) 2. Determines relative temporal resolutions 3. Applies appropriate interpolation: - Higher resolution aux → nearest neighbor - Lower resolution aux → specialized interpolator from metadata

Examples:

>>> from canvod.auxiliary.matching import DatasetMatcher
>>>
>>> matcher = DatasetMatcher()
>>> matched = matcher.match_datasets(
...     rinex_ds,
...     ephemerides=sp3_data,
...     clock=clk_data
... )
>>>
>>> # Auxiliary datasets now aligned to RINEX epochs
>>> len(matched['ephemerides'].epoch) == len(rinex_ds.epoch)
True
>>> len(matched['clock'].epoch) == len(rinex_ds.epoch)
True

Notes

Reference dataset should be the RINEX observations
Auxiliary datasets should have 'interpolator_config' in attrs
If no interpolator config, falls back to nearest neighbor
Temporal distance is tracked for quality assessment

`match_datasets(reference_ds, **aux_datasets)` ¶

Match auxiliary datasets to reference dataset epochs.

Parameters:

Name	Type	Description	Default
`reference_ds`	`Dataset`	Primary dataset (usually RINEX observations) that defines the target epoch timeline. This dataset remains unchanged.	required
`**aux_datasets`	`dict[str, Dataset]`	Named auxiliary datasets to align to reference epochs. Keys become the names in the returned dict. Example: ephemerides=sp3_data, clock=clk_data	`{}`

Returns:

Type	Description
`dict[str, Dataset]`	Dictionary of matched auxiliary datasets, all aligned to reference_ds.epoch. Keys match the input **aux_datasets keys.

Raises:

Type	Description
`ValueError`	If no auxiliary datasets provided If datasets missing required dimensions If interpolation config missing (warns, doesn't raise)

Examples:

>>> matcher = DatasetMatcher()
>>> matched = matcher.match_datasets(
...     rinex_ds,
...     ephemerides=sp3_data,
...     clock=clk_data
... )
>>> matched.keys()
dict_keys(['ephemerides', 'clock'])

`ECEFPosition` ¶

Bases: BaseModel

Earth-Centered, Earth-Fixed (ECEF) position in meters.

ECEF is a Cartesian coordinate system with: - Origin at Earth's center of mass - X-axis pointing to 0° latitude, 0° longitude (Prime Meridian at Equator) - Y-axis pointing to 0° latitude, 90° East longitude - Z-axis pointing to North Pole

Parameters:

Name	Type	Description	Default
`x`	`float`	X coordinate in meters.	required
`y`	`float`	Y coordinate in meters.	required
`z`	`float`	Z coordinate in meters.	required

Raises:

Type	Description
`ValidationError`	If any component is non-finite (NaN/inf), or the position is exactly (0, 0, 0) -- Earth's center, never a physically valid ECEF point for anything this codebase computes geometry relative to. This is deliberately a general sanity check (applies to satellite positions too, not just receivers) -- see `from_ds_metadata()` for the additional ground-station-specific plausibility check.

Examples:

>>> # From RINEX dataset metadata
>>> pos = ECEFPosition.from_ds_metadata(rinex_ds)
>>> print(f"X: {pos.x:.3f} m")

>>> # Manual creation
>>> pos = ECEFPosition(x=4194304.123, y=176481.234, z=4780013.456)
>>> lat, lon, alt = pos.to_geodetic()

`to_geodetic()` ¶

Convert ECEF to geodetic coordinates.

Returns:

Type	Description
`tuple[float, float, float]`	(latitude, longitude, altitude) where: - latitude: degrees [-90, 90] - longitude: degrees [-180, 180] - altitude: meters above WGS84 ellipsoid

`from_ds_metadata(ds)` `classmethod` ¶

Extract ECEF position from RINEX dataset metadata.

Reads from standard RINEX header attributes.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	RINEX dataset with position in attributes.	required

Returns:

Type	Description
`ECEFPosition`	Receiver position in ECEF.

Raises:

Type	Description
`KeyError`	If position attributes not found in dataset.
`ValidationError`	If the position is non-finite or exactly (0, 0, 0) (see `ECEFPosition`'s own validator).
`ValueError`	If the position's ECEF magnitude falls outside the plausible ground-station band (`_GROUND_STATION_MIN_RADIUS_M` .. `_GROUND_STATION_MAX_RADIUS_M`, i.e. on/near the WGS84 ellipsoid) -- e.g. an unpopulated RINEX header defaulted to a position that happens to be finite and non-zero but still not a real ground station (this check is specific to this parse boundary, not a general `ECEFPosition` constraint -- see the class docstring).

Examples:

>>> from canvod.readers import Rnxv3Obs
>>> rnx = Rnxv3Obs(fpath="station.24o")
>>> ds = rnx.to_ds()
>>> pos = ECEFPosition.from_ds_metadata(ds)

`GeodeticPosition` `dataclass` ¶

Geodetic (WGS84) position.

Parameters:

Name	Type	Description	Default
`lat`	`float`	Latitude in degrees [-90, 90].	required
`lon`	`float`	Longitude in degrees [-180, 180].	required
`alt`	`float`	Altitude in meters above WGS84 ellipsoid.	required

Examples:

>>> pos = GeodeticPosition(lat=48.208, lon=16.373, alt=200.0)
>>> print(f"Vienna: {pos.lat}°N, {pos.lon}°E, {pos.alt}m")

`to_ecef()` ¶

Convert geodetic to ECEF coordinates.

Returns:

Type	Description
`ECEFPosition`	Position in ECEF frame.

Preprocessing¶

Auxiliary-data preprocessing pipelines.

Composes the SID-space primitives from :mod:canvod.readers.preprocessing (sv->sid conversion, global SID padding, dtype/encoding normalization) into the two pipelines specific to auxiliary (SP3/CLK) data: preparing a dataset for Icechunk storage, and preparing one for interpolation.

The primitives themselves live in canvod-readers because they only depend on GNSS signal/constellation definitions (canvod.readers.gnss_specs), not on anything ephemeris- or interpolation-specific -- keeping them here made canvod-readers (which calls them from every reader's to_ds()) depend on canvod-auxiliary, which already depends on canvod-readers, a genuine import cycle that broke standalone canvod-readers installs.

`prep_aux_ds(aux_ds, fill_value=np.nan, aggregate_glonass_fdma=True, keep_sids=None)` ¶

Preprocess auxiliary dataset before writing to Icechunk.

Performs complete 4-step preprocessing: 1. Convert sv → sid dimension 2. Pad to global sid list (all constellations) or filter to keep_sids 3. Normalize sid dtype to object 4. Strip _FillValue attributes

Parameters:

Name	Type	Description	Default
`aux_ds`	`Dataset`	Dataset with 'sv' dimension.	required
`fill_value`	`float`	Fill value for missing entries.	`np.nan`
`aggregate_glonass_fdma`	`bool`	Whether to aggregate GLONASS FDMA bands.	`True`
`keep_sids`	`list[str] \| None`	List of specific SIDs to keep. If None, keeps all possible SIDs.	`None`

Returns:

Type	Description
`Dataset`	Fully preprocessed dataset ready for Icechunk or interpolation.

`preprocess_aux_for_interpolation(aux_ds, fill_value=np.nan, full_preprocessing=False, aggregate_glonass_fdma=True)` ¶

Preprocess auxiliary dataset before interpolation.

Converts satellite vehicle (sv) dimension to Signal ID (sid) dimension, which is required for matching with RINEX observations after interpolation.

Parameters:

Name	Type	Description	Default
`aux_ds`	`Dataset`	Raw auxiliary dataset with 'sv' dimension.	required
`fill_value`	`float`	Fill value for missing entries.	`np.nan`
`full_preprocessing`	`bool`	If True, applies full 4-step preprocessing (pad_to_global_sid, normalize_sid_dtype, strip_fillvalue). If False, only converts sv → sid (sufficient for interpolation).	`False`
`aggregate_glonass_fdma`	`bool`	Whether to aggregate GLONASS FDMA bands.	`True`

Returns:

Type	Description
`Dataset`	Preprocessed dataset with 'sid' dimension.

Notes

This must be called BEFORE interpolation. The workflow is: 1. Load raw SP3/CLK data (sv dimension) 2. Convert sv → sid (this function) 3. Interpolate to target epochs 4. Match with RINEX data (sid dimension)

For most interpolation use cases, full_preprocessing=False is sufficient. Use full_preprocessing=True when preparing data for Icechunk storage.

Examples:

>>> # Load raw SP3 data
>>> sp3_data = Sp3File(...).to_dataset()
>>> sp3_data.dims
{'epoch': 96, 'sv': 32}
>>>
>>> # Preprocess before interpolation (minimal)
>>> sp3_preprocessed = preprocess_aux_for_interpolation(sp3_data)
>>> sp3_preprocessed.dims
{'epoch': 96, 'sid': 384}
>>>
>>> # Preprocess before Icechunk (full)
>>> sp3_preprocessed = preprocess_aux_for_interpolation(
...     sp3_data,
...     full_preprocessing=True,
... )
>>> sp3_preprocessed.dims
{'epoch': 96, 'sid': ~2000}  # Padded to all possible sids
>>>
>>> # Now interpolate
>>> sp3_interp = interpolator.interpolate(sp3_preprocessed, target_epochs)

Interpolation¶

Interpolation strategies for GNSS auxiliary data.

`ClockConfig` ¶

Bases: InterpolatorConfig

Configuration for clock correction interpolation.

Attributes:

Name	Type	Description
`window_size`	`int, default 9`	Window size for discontinuity detection.
`jump_threshold`	`float, default 1e-6`	Threshold for detecting clock jumps (seconds).

`ClockInterpolationStrategy` ¶

Bases: Interpolator[ClockConfig]

Piecewise linear interpolation for clock corrections.

Detects and handles discontinuities (clock jumps) properly.

Examples:

>>> from canvod.auxiliary.interpolation import (
...     ClockInterpolationStrategy, ClockConfig,
... )
>>>
>>> config = ClockConfig(window_size=9, jump_threshold=1e-6)
>>> interpolator = ClockInterpolationStrategy(config=config)
>>>
>>> # Interpolate to RINEX epochs
>>> clk_interp = interpolator.interpolate(clk_data, rinex_epochs)

`interpolate(ds, target_epochs)` ¶

Interpolate clock corrections with discontinuity handling.

`Interpolator` ¶

Bases: ABC

Abstract base class for interpolation strategies.

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True and uses ABC to define required interpolation hooks.

`interpolate(ds, target_epochs)` `abstractmethod` ¶

Interpolate dataset to match target epochs.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Source dataset with (epoch, sid) dimensions.	required
`target_epochs`	`ndarray`	Target epoch grid (datetime64).	required

Returns:

Type	Description
`Dataset`	Interpolated dataset at target epochs.

`to_attrs()` ¶

Convert interpolator to attrs-compatible dictionary.

`InterpolatorConfig` ¶

Base class for interpolator configuration.

`to_dict()` ¶

Convert config to dictionary for attrs storage.

`Sp3Config` ¶

Bases: InterpolatorConfig

Configuration for SP3 ephemeris interpolation.

Attributes:

Name	Type	Description
`use_velocities`	`bool, default True`	Use Hermite splines with satellite velocities if available.
`fallback_method`	`str, default 'linear'`	Interpolation method when velocities are unavailable.

`Sp3InterpolationStrategy` ¶

Bases: Interpolator[Sp3Config]

Hermite cubic spline interpolation for SP3 ephemeris data.

Uses satellite velocities (Vx, Vy, Vz) for higher accuracy. Falls back to linear interpolation if velocities unavailable.

Examples:

>>> from canvod.auxiliary.interpolation import Sp3InterpolationStrategy, Sp3Config
>>>
>>> config = Sp3Config(use_velocities=True, fallback_method='linear')
>>> interpolator = Sp3InterpolationStrategy(config=config)
>>>
>>> # Interpolate to RINEX epochs
>>> sp3_interp = interpolator.interpolate(sp3_data, rinex_epochs)

`interpolate(ds, target_epochs)` ¶

Interpolate SP3 ephemeris to target epochs.

`aux_epoch_grid(day_start, interval_s)` ¶

Epoch grid of one day: day_start plus multiples of interval_s.

Parameters:

Name	Type	Description	Default
`day_start`	`datetime64`	00:00 of the day.	required
`interval_s`	`float`	Grid spacing in seconds (the sampling interval of the observations), resolved to the nanosecond.	required

Returns:

Type	Description
`ndarray`	`datetime64[ns]` epochs within the day.

Raises:

Type	Description
`ValueError`	If `interval_s` is not positive.

`interpolate_aux_day(ephem_ds, clock_ds, target_epochs)` ¶

Interpolate one day of orbits (and clocks) onto target_epochs.

Orbits: cubic Hermite spline with the satellite velocities (:class:Sp3InterpolationStrategy). Clocks: piecewise linear between clock jumps (:class:ClockInterpolationStrategy). Each part records its interpolator in attrs["interpolator_config"].

Parameters:

Name	Type	Description	Default
`ephem_ds`	`Dataset`	Orbits with `X`, `Y`, `Z` and `Vx`, `Vy`, `Vz`.	required
`clock_ds`	`Dataset \| None`	Clock offsets, or `None` when clocks are not fetched.	required
`target_epochs`	`ndarray`	Epoch grid, see :func:`aux_epoch_grid`.	required

Returns:

Type	Description
`Dataset`	Interpolated orbits, merged with the interpolated clocks if given.

`sampling_interval_from_epochs(epochs)` ¶

Sampling interval of observation epochs, in seconds.

The most common difference between consecutive epochs, so data gaps do not change the result.

Parameters:

Name	Type	Description	Default
`epochs`	`ndarray`	Observation epochs (datetime64), in time order.	required

Returns:

Type	Description
`float`	Sampling interval in seconds.

Raises:

Type	Description
`ValueError`	If there are fewer than two distinct epochs.

`create_interpolator_from_attrs(attrs)` ¶

Recreate interpolator instance from dataset attributes.

Parameters:

Name	Type	Description	Default
`attrs`	`dict`	Dataset attributes containing interpolator_config.	required

Returns:

Type	Description
`Interpolator`	Reconstructed interpolator instance.

Examples:

>>> # Save interpolator config in dataset
>>> ds.attrs['interpolator_config'] = interpolator.to_attrs()
>>>
>>> # Later, recreate interpolator
>>> interpolator = create_interpolator_from_attrs(ds.attrs)

Ephemeris (SP3)¶

Ephemeris (satellite orbit) data handling.

This module provides tools for reading, parsing, and validating satellite ephemeris data from SP3 format files, as well as the EphemerisProvider ABC for augmenting GNSS datasets with angular coordinates.

`Sp3Parser` ¶

Parser for SP3 (Standard Product #3) orbit files.

Handles parsing of SP3 format files containing precise satellite orbit data. Implements optimized single-pass reading for performance.

Parameters:

Name	Type	Description	Default
`fpath`	`Path`	Path to SP3 file.	required
`dimensionless`	`bool`	If True, strip units from output.	`True`

`init(fpath, dimensionless=True)` ¶

Initialize SP3 parser.

`parse()` ¶

Parse SP3 file to xarray Dataset.

Returns:

Type	Description
`Dataset`	Dataset with satellite positions (X, Y, Z) in meters.

Raises:

Type	Description
`FileNotFoundError`	If file does not exist.
`ValueError`	If file format is invalid.

`AgencyEphemerisProvider` ¶

Bases: EphemerisProvider

SP3/CLK-based ephemeris from analysis centers.

Downloads final or rapid products from COD, ESA, IGS etc., interpolates via Hermite cubic splines, and computes theta/phi via ECEF→spherical coordinate transformation.

Parameters:

Name	Type	Description	Default
`agency`	`str`	Analysis center code (`"COD"`, `"ESA"`, `"GFZ"`, `"JPL"`).	`'COD'`
`product_type`	`str`	Product type (`"final"` or `"rapid"`).	`'final'`
`aux_data_dir`	`Path`	Directory for cached aux Zarr files.	`None`
`keep_sids`	`list[str]`	SID filter list.	`None`
`store_radial_distance`	`bool`	Whether to keep `r` in the output.	`False`
`fetch_clock`	`bool`	Whether to also download/interpolate CLK clock-correction files. If None, uses `config.processing.aux_data.fetch_clock`.	`None`

`preprocess_day(date, site_config)` ¶

Download and read the SP3 (and CLK) products of one day.

The products are interpolated later, in :meth:augment_dataset, onto the day's epoch grid at the sampling interval of the dataset being augmented, the same grid canvodpy run uses (see :mod:canvod.auxiliary.interpolation.day_grid). The interpolated day is written once per sampling interval to aux_<date>_<interval>s.zarr in the returned directory.

Parameters:

Name	Type	Description	Default
`date`	`str`	Date in `YYYYDOY` format.	required
`site_config`	`Any`	Site configuration (needs `gnss_site_data_root`, receiver info).	required

Returns:

Type	Description
`Path`	Directory of the interpolated aux Zarr stores.

`augment_dataset(ds, receiver_position)` ¶

Add theta/phi to dataset using preprocessed SP3/CLK data.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	GNSS observation dataset.	required
`receiver_position`	`ECEFPosition`	Receiver ECEF position.	required

Returns:

Type	Description
`Dataset`	Augmented dataset with theta, phi (and optionally r).

Raises:

Type	Description
`RuntimeError`	If `preprocess_day()` has not been called.

`EphemerisProvider` ¶

Bases: ABC

Abstract base class for satellite ephemeris providers.

Implementations augment a GNSS dataset with angular coordinates (theta, phi) relative to a receiver position.

`augment_dataset(ds, receiver_position)` `abstractmethod` ¶

Add theta and phi (and optionally r) to ds.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	GNSS observation dataset with `(epoch, sid)` dims.	required
`receiver_position`	`ECEFPosition`	Receiver ECEF coordinates.	required

Returns:

Type	Description
`Dataset`	Dataset with `theta`, `phi` (and optionally `r`) added.

`preprocess_day(date, site_config)` `abstractmethod` ¶

Download / prepare ephemeris data for one day.

Parameters:

Name	Type	Description	Default
`date`	`str`	Date in `YYYYDOY` format.	required
`site_config`	`Any`	Site configuration object providing receiver info, data root, etc.	required

Returns:

Type	Description
`Path or None`	Path to cached/preprocessed ephemeris data, or `None` if preparation is not needed (e.g. broadcast provider).

`SbfBroadcastProvider` ¶

Bases: EphemerisProvider

Ephemeris from SBF SatVisibility broadcast data.

Transfers theta/phi directly from SBF metadata datasets, skipping external orbit/clock downloads entirely. Only works when the source format is SBF.

Only angles the receiver computed from the broadcast ephemeris are used (broadcast_angle_source == 2). Almanac-based angles become NaN: on the canVODpy test data (2025-001) they differ from the ephemeris-based angles by up to 0.26 deg for GLONASS and up to 104 deg for BeiDou.

Parameters:

Name	Type	Description	Default
`canopy_file`	`Path`	Path to canopy SBF file for reference receivers in shared-position mode. When provided, the canopy file's geometry overrides the reference file's own geometry.	`None`
`canopy_reader_format`	`str`	Reader format for the canopy file. Only `"sbf"` carries SatVisibility geometry.	`'sbf'`
`keep_sids`	`list of str`	SIDs to keep when reading `canopy_file`. Only saves work: the geometry is aligned to the observation SIDs either way.	`None`

`preprocess_day(date, site_config)` ¶

No preprocessing needed for broadcast ephemeris.

`augment_dataset(ds, receiver_position, *, aux_datasets=None)` ¶

Add theta/phi from SBF SatVisibility metadata.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	SBF observation dataset.	required
`receiver_position`	`ECEFPosition`	Receiver ECEF position (unused but required by ABC).	required
`aux_datasets`	`dict`	Auxiliary datasets from `reader.to_ds_and_auxiliary()`. Must contain `"sbf_obs"` with theta/phi.	`None`

Returns:

Type	Description
`Dataset`	Dataset with theta/phi added from SBF broadcast. Cells without an ephemeris-based angle are NaN.

Raises:

Type	Description
`ValueError`	If no `sbf_obs` is available, it lacks the SatVisibility variables, or it holds no ephemeris-based angle for any observation.

`Sp3File` ¶

Bases: AuxFile

Handler for SP3 orbit files with multi-product support.

Now supports all IGS analysis centers via product registry: COD, GFZ, ESA, JPL, IGS, WHU, GRG, SHA

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True.

Attributes:

Name	Type	Description
`date`	`str`	String in YYYYDOY format.
`agency`	`str`	Agency code (e.g., "COD", "GFZ", "ESA").
`product_type`	`str`	Product type ("final", "rapid").
`ftp_server`	`str`	Base URL for downloads.
`local_dir`	`Path`	Local storage directory.
`add_velocities`	`bool \| None, default True`	Whether to compute velocities.
`dimensionless`	`bool \| None, default True`	Whether to strip units (store magnitudes only).
`product_spec`	`(ProductSpec \| None, optional)`	Product specification resolved from the registry.

`__post_init__()` ¶

Initialize with product validation.

`get_interpolation_strategy()` ¶

Get appropriate interpolation strategy for SP3 files.

`candidate_filenames()` ¶

SP3 file names, one per sampling the product has been published with.

Pattern: {PREFIX}{YYYYDOY}0000_01D_ORB.SP3

Example: COD0MGXFIN_20240150000_01D_05M_ORB.SP3

`download_aux_file(filename)` ¶

Download one SP3 file, trying all servers from the product registry.

Servers are tried in priority order from products.toml. Auth-required servers are skipped when no credentials are configured. Warnings are printed when falling back to alternate servers.

Parameters:

Name	Type	Description	Default
`filename`	`str`	One of :meth:`candidate_filenames`.	required

Raises:

Type	Description
`RuntimeError`	If download fails from all available servers.
`ValueError`	If GPS week calculation fails.

`read_file()` ¶

Read and validate SP3 file.

Returns:

Type	Description
`Dataset`	Dataset with satellite positions (X, Y, Z) in meters.

Raises:

Type	Description
`FileNotFoundError`	If file does not exist.
`ValueError`	If validation fails.

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

Like :meth:AuxFile.join_days, with the velocities recomputed.

The velocities of each file have one-sided differences at its first and last epoch; recomputed on the joined data, the day's first and last epochs get central differences.

`compute_velocity(ds)` ¶

Compute satellite velocities from position data.

Uses central differences for interior points, forward/backward differences for endpoints, with the actual epoch spacing (the days joined by :meth:join_days may come with different samplings).

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset with X, Y, Z coordinates.	required

Returns:

Type	Description
`Dataset`	Dataset augmented (or updated) with Vx, Vy, Vz velocities.

`Sp3Validator` ¶

Validator for SP3 orbit files.

Performs format and data quality checks on parsed SP3 datasets.

Parameters:

Name	Type	Description	Default
`dataset`	`Dataset`	Parsed SP3 dataset.	required
`fpath`	`Path`	Path to the original file.	required

`init(dataset, fpath)` ¶

Initialize validator.

`validate()` ¶

Run all validation checks.

Returns:

Type	Description
`FileValidationResult`	Validation result with errors and warnings.

`get_summary()` ¶

Get validation summary.

Clock (CLK)¶

Optional: only when aux_data.fetch_clock is true (the default is false); the VOD formula doesn't use them.

Clock correction data handling.

This module provides tools for reading, parsing, and validating satellite clock correction data from RINEX CLK format files.

`ClkFile` ¶

Bases: AuxFile

Handler for GNSS clock files in CLK format.

This class reads and processes clock offset files containing satellite clock corrections. It handles the parsing of CLK format files and provides the data in xarray Dataset format.

Supports multiple analysis centers via product registry with proper FTP paths and filename conventions.

Notes

This is a Pydantic dataclass with arbitrary_types_allowed=True.

Attributes:

Name	Type	Description
`date`	`str`	String in YYYYDOY format representing the start date.
`agency`	`str`	Analysis center identifier (e.g., "COD", "GFZ").
`product_type`	`str`	Product type ("final" or "rapid").
`ftp_server`	`str`	Base URL for file downloads.
`local_dir`	`Path`	Local storage directory.
`dimensionless`	`bool \| None, default True`	If True, outputs magnitude-only values (no units attached).

`__post_init__()` ¶

Initialize CLK file handler.

`get_interpolation_strategy()` ¶

Get appropriate interpolation strategy for CLK files.

`candidate_filenames()` ¶

CLK file names, one per sampling the product has been published with.

Filename format: {PREFIX}{YYYYDOY}0000_01D_CLK.CLK

Raises:

Type	Description
`UnknownProductError`	If agency/product combination not in registry.
`ValueError`	If the product publishes no clock files.

`download_aux_file(filename)` ¶

Download one CLK file, trying all servers from the product registry.

Servers are tried in priority order from products.toml. Auth-required servers are skipped when no credentials are configured. Warnings are printed when falling back to alternate servers.

Parameters:

Name	Type	Description	Default
`filename`	`str`	One of :meth:`candidate_filenames`.	required

Raises:

Type	Description
`RuntimeError`	If file cannot be downloaded from any available server.
`ValueError`	If GPS week calculation fails.

`read_file()` ¶

Read and parse CLK file into xarray Dataset.

Uses modular parser for data extraction and validator for quality checks. The clock bias is in seconds (RINEX clock 3.04, Table A16).

Returns:

Type	Description
`Dataset`	Clock offsets with dimensions (epoch, sv). Values are in seconds (or dimensionless if specified).

`parse_clk_data(file_handle)` ¶

Parse CLK data records using two-pass strategy.

Two-Pass Strategy: Pass 1: Collect all unique epochs and satellites Pass 2: Fill data arrays with clock offsets

This ensures we capture all satellites, even if they're not in the header, and handles satellites appearing/disappearing during the time period.

Parameters:

Name	Type	Description	Default
`file_handle`	`TextIO`	Open file object positioned after header.	required

Returns:

Type	Description
`tuple[list[datetime], list[str], ndarray]`	(epochs, satellites, clock_offsets) where: - epochs: list of datetime objects - satellites: sorted list of satellite codes - clock_offsets: 2D array (epochs × satellites) in seconds

`parse_clk_file(filepath)` ¶

Parse complete CLK file.

Parameters:

Name	Type	Description	Default
`filepath`	`Path`	Path to CLK file.	required

Returns:

Type	Description
`tuple[list[datetime], list[str], ndarray]`	(epochs, satellites, clock_offsets).

`parse_clk_header(file_handle)` ¶

Parse CLK file header to extract satellite list.

Parameters:

Name	Type	Description	Default
`file_handle`	`TextIO`	Open file object positioned at start.	required

Returns:

Type	Description
`set[str]`	Satellite identifiers (e.g., "G01", "R01").

Raises:

Type	Description
`ValueError`	If header format is invalid, or if the time system (`TIME SYSTEM ID`; GPS if absent) is not GPS time.

`check_clk_data_quality(ds, min_coverage=80.0)` ¶

Check if CLK data meets minimum quality requirements.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset from a CLK file.	required
`min_coverage`	`float`	Minimum required data coverage percentage.	`80.0`

Returns:

Type	Description
`bool`	True if data quality is acceptable.

`validate_clk_dataset(ds)` ¶

Validate CLK dataset structure and data quality.

Checks: - Required variable exists (clock_offset) - Required coordinates exist (epoch, sv) - Data completeness (percentage of valid values) - Temporal consistency (monotonic epochs)

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset from a CLK file.	required

Returns:

Type	Description
`dict[str, bool \| float \| int]`	Validation results with keys: - has_clock_offset - has_epoch - has_sv - valid_data_percent - epochs_monotonic - num_epochs - num_satellites

Position and coordinates¶

Position and coordinate transformations for GNSS data.

Provides ECEF/geodetic position representations and spherical coordinate computation for satellite-receiver geometry analysis.

`ECEFPosition` ¶

Bases: BaseModel

Earth-Centered, Earth-Fixed (ECEF) position in meters.

ECEF is a Cartesian coordinate system with: - Origin at Earth's center of mass - X-axis pointing to 0° latitude, 0° longitude (Prime Meridian at Equator) - Y-axis pointing to 0° latitude, 90° East longitude - Z-axis pointing to North Pole

Parameters:

Name	Type	Description	Default
`x`	`float`	X coordinate in meters.	required
`y`	`float`	Y coordinate in meters.	required
`z`	`float`	Z coordinate in meters.	required

Raises:

Type	Description
`ValidationError`	If any component is non-finite (NaN/inf), or the position is exactly (0, 0, 0) -- Earth's center, never a physically valid ECEF point for anything this codebase computes geometry relative to. This is deliberately a general sanity check (applies to satellite positions too, not just receivers) -- see `from_ds_metadata()` for the additional ground-station-specific plausibility check.

Examples:

>>> # From RINEX dataset metadata
>>> pos = ECEFPosition.from_ds_metadata(rinex_ds)
>>> print(f"X: {pos.x:.3f} m")

>>> # Manual creation
>>> pos = ECEFPosition(x=4194304.123, y=176481.234, z=4780013.456)
>>> lat, lon, alt = pos.to_geodetic()

`to_geodetic()` ¶

Convert ECEF to geodetic coordinates.

Returns:

Type	Description
`tuple[float, float, float]`	(latitude, longitude, altitude) where: - latitude: degrees [-90, 90] - longitude: degrees [-180, 180] - altitude: meters above WGS84 ellipsoid

`from_ds_metadata(ds)` `classmethod` ¶

Extract ECEF position from RINEX dataset metadata.

Reads from standard RINEX header attributes.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	RINEX dataset with position in attributes.	required

Returns:

Type	Description
`ECEFPosition`	Receiver position in ECEF.

Raises:

Type	Description
`KeyError`	If position attributes not found in dataset.
`ValidationError`	If the position is non-finite or exactly (0, 0, 0) (see `ECEFPosition`'s own validator).
`ValueError`	If the position's ECEF magnitude falls outside the plausible ground-station band (`_GROUND_STATION_MIN_RADIUS_M` .. `_GROUND_STATION_MAX_RADIUS_M`, i.e. on/near the WGS84 ellipsoid) -- e.g. an unpopulated RINEX header defaulted to a position that happens to be finite and non-zero but still not a real ground station (this check is specific to this parse boundary, not a general `ECEFPosition` constraint -- see the class docstring).

Examples:

>>> from canvod.readers import Rnxv3Obs
>>> rnx = Rnxv3Obs(fpath="station.24o")
>>> ds = rnx.to_ds()
>>> pos = ECEFPosition.from_ds_metadata(ds)

`GeodeticPosition` `dataclass` ¶

Geodetic (WGS84) position.

Parameters:

Name	Type	Description	Default
`lat`	`float`	Latitude in degrees [-90, 90].	required
`lon`	`float`	Longitude in degrees [-180, 180].	required
`alt`	`float`	Altitude in meters above WGS84 ellipsoid.	required

Examples:

>>> pos = GeodeticPosition(lat=48.208, lon=16.373, alt=200.0)
>>> print(f"Vienna: {pos.lat}°N, {pos.lon}°E, {pos.alt}m")

`to_ecef()` ¶

Convert geodetic to ECEF coordinates.

Returns:

Type	Description
`ECEFPosition`	Position in ECEF frame.

`add_broadcast_spherical_coords_to_dataset(ds, theta, phi)` ¶

Add broadcast spherical coordinates to xarray Dataset.

Same convention as :func:add_spherical_coords_to_dataset but attrs reflect that values come from the receiver's SBF SatVisibility block (broadcast ephemeris) rather than independently-computed SP3/CLK ephemerides.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset with 'epoch' and 'sid' dimensions.	required
`theta`	`ndarray`	Polar angles in radians [0, π].	required
`phi`	`ndarray`	Azimuthal angles in radians [0, 2π).	required

Returns:

Type	Description
`Dataset`	Dataset with phi and theta variables added.

`add_spherical_coords_to_dataset(ds, r, theta, phi)` ¶

Add spherical coordinates to xarray Dataset with proper metadata.

Parameters:

Name	Type	Description	Default
`ds`	`Dataset`	Dataset with 'epoch' and 'sid' dimensions.	required
`r`	`ndarray`	Radial distances in meters.	required
`theta`	`ndarray`	Polar angles in radians [0, π].	required
`phi`	`ndarray`	Azimuthal angles in radians [0, 2π).	required

Returns:

Type	Description
`Dataset`	Dataset with phi, theta, r variables added.

Notes

Variables are added with CF-compliant attributes following physics convention.

Examples:

>>> # After computing spherical coordinates
>>> r, theta, phi = compute_spherical_coordinates(sat_x, sat_y, sat_z, rx_pos)
>>>
>>> # Add to RINEX dataset
>>> augmented_ds = add_spherical_coords_to_dataset(rinex_ds, r, theta, phi)
>>> print(augmented_ds.phi.attrs['description'])

`compute_spherical_coordinates(sat_x, sat_y, sat_z, rx_pos)` ¶

Compute spherical coordinates (r, theta, phi) in navigation convention.

Uses local ENU (East-North-Up) topocentric frame centered at receiver.

Navigation Convention: - theta: Polar angle from +z axis (zenith), [0, π] radians * theta = 0 → zenith (straight up) * theta = π/2 → horizon * theta > π/2 → below horizon (set to NaN) - phi: Azimuthal angle from North, clockwise, [0, 2π) radians * phi = 0 → North * phi = π/2 → East * phi = π → South * phi = 3π/2 → West - r: Radial distance in meters

Note

The phi convention follows geographic/navigation standards where: - 0° points North (positive Y in ENU frame) - Angles increase clockwise when viewed from above - This is computed as arctan2(East, North), giving North=0° reference - Differs from physics convention which uses East=0° reference

Parameters:

Name	Type	Description	Default
`sat_x`	`ndarray`	Satellite X coordinates in ECEF (meters).	required
`sat_y`	`ndarray`	Satellite Y coordinates in ECEF (meters).	required
`sat_z`	`ndarray`	Satellite Z coordinates in ECEF (meters).	required
`rx_pos`	`ECEFPosition`	Receiver position in ECEF.	required

Returns:

Type	Description
`tuple[ndarray, ndarray, ndarray]`	(r, theta, phi) where: - r: distances in meters - theta: polar angles in radians [0, π] - phi: azimuthal angles in radians [0, 2π)

Notes

Satellites below horizon (theta > π/2) are set to NaN.

Examples:

>>> from canvod.auxiliary.position import (
...     ECEFPosition, compute_spherical_coordinates,
... )
>>>
>>> # Receiver position
>>> rx = ECEFPosition(x=4194304.0, y=176481.0, z=4780013.0)
>>>
>>> # Satellite positions (example)
>>> sat_x = np.array([16364123.0, 10205789.0])
>>> sat_y = np.array([12123456.0, -8901234.0])
>>> sat_z = np.array([18456789.0, 21234567.0])
>>>
>>> r, theta, phi = compute_spherical_coordinates(sat_x, sat_y, sat_z, rx)
>>> print(f"Distance: {r[0]/1e6:.2f} Mm")
>>> print(f"Polar angle: {np.degrees(theta[0]):.1f}°")
>>> print(f"Azimuth: {np.degrees(phi[0]):.1f}°")

Product registry¶

Product registry and specifications for IGS analysis centers.

`ProductSpec` ¶

Bases: BaseModel

Product specification with structural validation.

A product is published as one SP3 and, if it has clocks, one CLK file per day. Agencies have changed the sampling of their files over the years, so sp3_samplings and clk_samplings list every sampling the product has been published with, in the order they are tried.

Validates data structure only (fast). FTP availability is checked during download, and for the whole registry by python -m canvod.auxiliary.products.check (just check-products).

Notes

This is a Pydantic BaseModel.

`has_clock` `property` ¶

Whether the product publishes CLK files.

`available_formats` `property` ¶

File formats the product publishes.

`validate_samplings(v)` `classmethod` ¶

Validate sampling codes such as 05M or 30S.

`validate_prefix_matches_agency(v, info)` `classmethod` ¶

Validate prefix starts with agency code.

`sp3_filenames(yyyydoy)` ¶

SP3 file names for one day, in the order they are tried.

Parameters:

Name	Type	Description	Default
`yyyydoy`	`str`	Day as `YYYYDOY`.	required

Returns:

Type	Description
`list[str]`	One name per sampling, without the `.gz` of the server copy.

`clk_filenames(yyyydoy)` ¶

CLK file names for one day, in the order they are tried.

Parameters:

Name	Type	Description	Default
`yyyydoy`	`str`	Day as `YYYYDOY`.	required

Returns:

Type	Description
`list[str]`	One name per sampling; empty if the product has no clocks.

`get_product_spec(agency, product_type)` ¶

Get product spec from global registry.

`list_agencies()` ¶

List agencies from global registry.

`get_products_for_agency(agency)` ¶

Get products for agency from global registry.

canvod.auxiliary API reference¶

Package¶

Sp3File ¶

__post_init__() ¶

get_interpolation_strategy() ¶

candidate_filenames() ¶

download_aux_file(filename) ¶

read_file() ¶

join_days(previous, following, margin=DAY_JOIN_MARGIN) ¶

compute_velocity(ds) ¶

ClkFile ¶

__post_init__() ¶

get_interpolation_strategy() ¶

candidate_filenames() ¶

download_aux_file(filename) ¶

read_file() ¶

AuxFile ¶

data property ¶

__post_init__() ¶

from_datetime_date(date, agency, product_type, ftp_server, local_dir, **kwargs) classmethod ¶

from_file(fpath, **kwargs) classmethod ¶

download_file(url, destination, file_info=None) ¶

download_with_fallback(ftp_path, destination, file_info=None, product_spec=None) ¶

read_file() abstractmethod ¶

get_interpolation_strategy() abstractmethod ¶

join_days(previous, following, margin=DAY_JOIN_MARGIN) ¶

check_file_exists() ¶

candidate_filenames() abstractmethod ¶

download_aux_file(filename) abstractmethod ¶

Interpolator ¶

interpolate(ds, target_epochs) abstractmethod ¶

to_attrs() ¶

InterpolatorConfig ¶

to_dict() ¶

DatasetMatcher ¶

match_datasets(reference_ds, **aux_datasets) ¶

ECEFPosition ¶

to_geodetic() ¶

from_ds_metadata(ds) classmethod ¶

GeodeticPosition dataclass ¶

to_ecef() ¶

Preprocessing¶

prep_aux_ds(aux_ds, fill_value=np.nan, aggregate_glonass_fdma=True, keep_sids=None) ¶

preprocess_aux_for_interpolation(aux_ds, fill_value=np.nan, full_preprocessing=False, aggregate_glonass_fdma=True) ¶

Interpolation¶

ClockConfig ¶

ClockInterpolationStrategy ¶

interpolate(ds, target_epochs) ¶

Interpolator ¶

interpolate(ds, target_epochs) abstractmethod ¶

to_attrs() ¶

InterpolatorConfig ¶

to_dict() ¶

Sp3Config ¶

Sp3InterpolationStrategy ¶

interpolate(ds, target_epochs) ¶

aux_epoch_grid(day_start, interval_s) ¶

interpolate_aux_day(ephem_ds, clock_ds, target_epochs) ¶

sampling_interval_from_epochs(epochs) ¶

create_interpolator_from_attrs(attrs) ¶

Ephemeris (SP3)¶

Sp3Parser ¶

__init__(fpath, dimensionless=True) ¶

parse() ¶

AgencyEphemerisProvider ¶

preprocess_day(date, site_config) ¶

augment_dataset(ds, receiver_position) ¶

EphemerisProvider ¶

augment_dataset(ds, receiver_position) abstractmethod ¶

preprocess_day(date, site_config) abstractmethod ¶

SbfBroadcastProvider ¶

preprocess_day(date, site_config) ¶

augment_dataset(ds, receiver_position, *, aux_datasets=None) ¶

Sp3File ¶

__post_init__() ¶

get_interpolation_strategy() ¶

candidate_filenames() ¶

download_aux_file(filename) ¶

read_file() ¶

join_days(previous, following, margin=DAY_JOIN_MARGIN) ¶

`canvod.auxiliary` API reference¶

`Sp3File` ¶

`__post_init__()` ¶

`get_interpolation_strategy()` ¶

`candidate_filenames()` ¶

`download_aux_file(filename)` ¶

`read_file()` ¶

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

`compute_velocity(ds)` ¶

`ClkFile` ¶

`__post_init__()` ¶

`get_interpolation_strategy()` ¶

`candidate_filenames()` ¶

`download_aux_file(filename)` ¶

`read_file()` ¶

`AuxFile` ¶

`data` `property` ¶

`__post_init__()` ¶

`from_datetime_date(date, agency, product_type, ftp_server, local_dir, **kwargs)` `classmethod` ¶

`from_file(fpath, **kwargs)` `classmethod` ¶

`download_file(url, destination, file_info=None)` ¶

`download_with_fallback(ftp_path, destination, file_info=None, product_spec=None)` ¶

`read_file()` `abstractmethod` ¶

`get_interpolation_strategy()` `abstractmethod` ¶

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

`check_file_exists()` ¶

`candidate_filenames()` `abstractmethod` ¶

`download_aux_file(filename)` `abstractmethod` ¶

`Interpolator` ¶

`interpolate(ds, target_epochs)` `abstractmethod` ¶

`to_attrs()` ¶

`InterpolatorConfig` ¶

`to_dict()` ¶

`DatasetMatcher` ¶

`match_datasets(reference_ds, **aux_datasets)` ¶

`ECEFPosition` ¶

`to_geodetic()` ¶

`from_ds_metadata(ds)` `classmethod` ¶

`GeodeticPosition` `dataclass` ¶

`to_ecef()` ¶

`prep_aux_ds(aux_ds, fill_value=np.nan, aggregate_glonass_fdma=True, keep_sids=None)` ¶

`preprocess_aux_for_interpolation(aux_ds, fill_value=np.nan, full_preprocessing=False, aggregate_glonass_fdma=True)` ¶

`ClockConfig` ¶

`ClockInterpolationStrategy` ¶

`interpolate(ds, target_epochs)` ¶

`Interpolator` ¶

`interpolate(ds, target_epochs)` `abstractmethod` ¶

`to_attrs()` ¶

`InterpolatorConfig` ¶

`to_dict()` ¶

`Sp3Config` ¶

`Sp3InterpolationStrategy` ¶

`interpolate(ds, target_epochs)` ¶

`aux_epoch_grid(day_start, interval_s)` ¶

`interpolate_aux_day(ephem_ds, clock_ds, target_epochs)` ¶

`sampling_interval_from_epochs(epochs)` ¶

`create_interpolator_from_attrs(attrs)` ¶

`Sp3Parser` ¶

`init(fpath, dimensionless=True)` ¶

`parse()` ¶

`AgencyEphemerisProvider` ¶

`preprocess_day(date, site_config)` ¶

`augment_dataset(ds, receiver_position)` ¶

`EphemerisProvider` ¶

`augment_dataset(ds, receiver_position)` `abstractmethod` ¶

`preprocess_day(date, site_config)` `abstractmethod` ¶

`SbfBroadcastProvider` ¶

`preprocess_day(date, site_config)` ¶

`augment_dataset(ds, receiver_position, *, aux_datasets=None)` ¶

`Sp3File` ¶

`__post_init__()` ¶

`get_interpolation_strategy()` ¶

`candidate_filenames()` ¶

`download_aux_file(filename)` ¶

`read_file()` ¶

`join_days(previous, following, margin=DAY_JOIN_MARGIN)` ¶

`compute_velocity(ds)` ¶

`Sp3Validator` ¶

`init(dataset, fpath)` ¶

`validate()` ¶