osdyn.utils.data.select

Created on Mon Oct 26 21:10:23 2020

@author: valerie

Module Contents

Functions

modify_where(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)

mask_where(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)

Mask the elements satisfying the condition.

select_where_mask_elsewhere(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)

Select the elements by masking all of those

select_where_mask_elsewhere_from_np(npa, flag={'automatic': {'test': '==', 'value': 1}}, index_condition_field=None, **kwargs)

Apply the filter (defined by flag) on the variables specified in the

select_where_mask_elsewhere_from_ds(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}}, outformat='list')

Apply the filter (defined by flag) on the variables specified by varnames

flag2cmd(flag)

Return the command to be applied on a field

test_mask_where_np(varname='TEMP_ADJUSTED')

Attributes

opposite

osdyn.utils.data.select.opposite[source]
osdyn.utils.data.select.modify_where(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)[source]
osdyn.utils.data.select.mask_where(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)[source]

Mask the elements satisfying the condition.

As it uses select_where_mask_elsewhere function, the first step is to modify the condition to get its opposite

Parameters

  • ds (xarray.Dataset, numpy.ndarray, numpy.array, list (of numpy.array)) – Dataset or arrays containing the fields to be modified.

  • varnames ([list of] str, optional) – Names of the variables of interest. The default is None, then all the variables of the dataset are treated.

  • flag (dict, optional) – {“VARNAME”: {“test”: “operator >, ==…”, “value”: X}}. The default is {“POSITION_QC”: {“test”: “==”, “value”: 1}, “fromds”: “ds”}.

  • outformat (str, optional) – ‘dataset’ or ‘list’. The default is ‘list’.

  • kwargs (index_condition_field in) – Location (from 0) of the field (in ds) used in the condition. Usefull when processing numpy arrays. The default is None.

Returns

sel – that contains the masked variables according to the flag.

Return type

dataset or list of arrays

osdyn.utils.data.select.select_where_mask_elsewhere(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}, 'fromds': 'ds'}, outformat='list', **kwargs)[source]

Select the elements by masking all of those which to not satisfy the condition.

Parameters

  • ds (xarray.Dataset, numpy.ndarray, numpy.array, list (of numpy.array)) – Dataset or arrays containing the fields to be modified.

  • varnames ([list of] str, optional) – Names of the variables of interest. The default is None, then all the variables of the dataset are treated.

  • flag (dict, optional) – {“VARNAME”: {“test”: “operator >, ==…”, “value”: X}}. The default is {“POSITION_QC”: {“test”: “==”, “value”: 1}, “fromds”: “ds”}.

  • outformat (str, optional) – ‘dataset’ or ‘list’. The default is ‘list’.

  • kwargs (index_condition_field in) – Location (from 0) of the field (in ds) used in the condition. Usefull when processing numpy arrays. The default is None.

Returns

sel – that contains the masked variables according to the flag.

Return type

dataset or list of arrays

Todo

add argument ‘other’ (for modify_where function)

osdyn.utils.data.select.select_where_mask_elsewhere_from_np(npa, flag={'automatic': {'test': '==', 'value': 1}}, index_condition_field=None, **kwargs)[source]

Apply the filter (defined by flag) on the variables specified in the first argument (npa) to keep values where the condition is satisfied (others are set to np.nan).

Parameters

  • npa ([list of] np.ndarray) – Fields to be masked or modified.

  • flag (dict, optional) – for instance {“whatever”: {“test”: “>”, “value”: 111.}. The default is {“automatic”: {“test”: “==”, “value”: 1}}.

  • index_condition_field (int, optional) – Location (from 0) of the field (in npa) used in the condition. The default is None.

Returns

sel – that contain the masked variables according to the flag.

Return type

list of numpy arrays

osdyn.utils.data.select.select_where_mask_elsewhere_from_ds(ds, varnames=None, flag={'POSITION_QC': {'test': '==', 'value': 1}}, outformat='list')[source]

Apply the filter (defined by flag) on the variables specified by varnames to keep values where the condition is satisfied (others are set to np.nan).

Parameters

  • ds (xarray.Dataset) – Content of a dataset.

  • varnames ([list of] str, optional) – Names of the variables of interest. The default is None, then all the variables are treated.

  • flag (dict, optional) – {“VARNAME”: {“test”: “operator >, ==…”, “value”: X}}. The default is {“POSITION_QC”: {“test”: “==”, “value”: 1}, “fromds”: “ds”}.

  • outformat (str, optional) – ‘dataset’ or ‘list’. The default is ‘list’.

Returns

sel – that contains the masked variables according to the flag.

Return type

dataset, list

osdyn.utils.data.select.flag2cmd(flag)[source]

Return the command to be applied on a field

Parameters

flag (dict) –

  • dict like {‘POSITION_QC’: {“test”: “==”, “value”: 1}} or {‘POSITION_QC’: 1}

Returns

comparison command to be applied on data.

Return type

str

Examples

>>> flag2cmd({"POSITION_QC": {"test": "==", "value": 1}})
>>> flag2cmd({'POSITION_QC': 1, "fromds": "ds"})
osdyn.utils.data.select.test_mask_where_np(varname='TEMP_ADJUSTED')[source]