Dataset¶

class spy4cast.dataset.Dataset(name: str, folder: str = '.', chunks: Any | None = None)¶

Bases: object

This class enables you to load, slice and save the data in a netcdf4 file confortably.

It is concatenable, meaning that the output of most methods is the object itself so that you can concatenate methods easily.

Parameters:

name (str, default='') – Directory where the dataset you want to use is located
folder (str, default='dataset.nc') – Name of the dataset
chunks (xarray.core.types.T_Chunks, optional) – Argument passed when loading the datasets (see chunks in dask library)

Example

>>> Dataset(
...     'example.nc', 'data/'
... ).open('var').slice(
...     Region(-20, 20, -10, 0, Month.JAN, Month.FEB, 1870, 2000)
... ).save_nc('data-sliced.nc')

Attributes Summary

`data`	Returns the data contained
`lat`	Returns the latitude variable of the data evaluated.
`lon`	Returns the latitude variable of the data evaluated.
`region`	Returns the actual region of the data
`shape`	Returns the shape variable of the data evaluated.
`time`	Returns the time variable of the data evaluated.
`timestamp0`	Initial timestamp of the data
`timestampf`	Final timestamp of the data
`var`	Return the variable used.

Methods Summary

`from_xrarray`(array)	Create a dataset from a xarray
`open`([var, region])	Opens dataset without loading it into memory
`save_nc`(name[, folder])	Saves the data as a netcdf4 file
`slice`(region[, skip])	Method that slices the dataset accorging to a Region.

Attributes Documentation

data¶: Returns the data contained

lat¶: Returns the latitude variable of the data evaluated. They key used is recognised automatically

lon¶: Returns the latitude variable of the data evaluated. They key used is recognised automatically

region¶: Returns the actual region of the data

shape¶: Returns the shape variable of the data evaluated.

time¶: Returns the time variable of the data evaluated. They key used is recognised automatically

timestamp0¶: Initial timestamp of the data

timestampf¶: Final timestamp of the data

var¶

Return the variable used.

Raises:: ValueError – If the dataset has not been opened yet

Methods Documentation

classmethod from_xrarray(array: DataArray) → Dataset¶: Create a dataset from a xarray

open(var: str | None = None, region: Region | None = None) → _T¶

Opens dataset without loading it into memory

Warning

If the dataset says that the dataset starts from a year smaller than 1678 (minimum for pandas), it will be loaded with initial year = 2000

Parameters:

var (optional, str) – Variable that can be specified. If not it is detected automatically
region (optional, Region) – Fill in this field the dataset is already sliced and you want to specify a certain region

Raises:

errors.DatasetError – If there is an error while opening the dataset
errors.DatasetNotFoundError – If the dataset name or folder is not valid
errors.VariableSelectionError – If teh variable selected does not exist or can not be inferred

save_nc(name: str, folder: str = '.') → _T¶: Saves the data as a netcdf4 file

slice(region: Region, skip: int = 0) → _T¶

Method that slices the dataset accorging to a Region.

Parameters:

region (spy4cast.stypes.Region) – spy4cast.stypes.Region to use
skip (int) – Amount of points to skip in the matrix

Note

It first calls check_region method

Note

If the season contains months from different years (NOV-DEC-JAN-FEB for example) the initial year is applied to the month which comes at last (FEB). In this example, the data that will be used for NOV is on year before the initial year so keep this in mind if your dataset doesn’t contain that specific year.