From a74fdc4055993a17b80c567d175795991b2bf5f9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Hampus=20N=C3=A4sstr=C3=B6m?= <hampus.nasstrom@gmail.com>
Date: Wed, 22 Feb 2023 16:37:24 +0100
Subject: [PATCH] Added Function for Generating NumericalArray Class
---
nomad/datamodel/metainfo/eln/__init__.py | 132 ++++++++++++++++++++++-
1 file changed, 131 insertions(+), 1 deletion(-)
diff --git a/nomad/datamodel/metainfo/eln/__init__.py b/nomad/datamodel/metainfo/eln/__init__.py
index e6dfc4f011..df1102916e 100644
--- a/nomad/datamodel/metainfo/eln/__init__.py
+++ b/nomad/datamodel/metainfo/eln/__init__.py
@@ -19,7 +19,7 @@
import numpy as np
import datetime
import re
-from typing import Any, Dict, List
+from typing import Any, Dict, List, Type, Union
from nomad import utils
from nomad.units import ureg
from nomad.datamodel.data import EntryData, ArchiveSection, author_reference, BasicElnCategory
@@ -40,6 +40,136 @@ from nomad.datamodel.metainfo.eln.nexus_data_converter import (
m_package = Package(name='eln')
+def numerical_array(unit: str = None, shape: List[Union[str, int]] = None,
+ dimensionality: int = None) -> Type[ArchiveSection]:
+ '''
+ Function for generating a NumericalArray section with the specified shape and units.
+
+ Args:
+ unit (str, optional): The units of the values in the array. Defaults to None.
+ shape (List[Union[str, int]], optional): The shape of the array. Defaults to None.
+ dimensionality (int, optional): The dimensionality of the array. Defaults to None.
+
+ Raises:
+ ValueError: If both shape and dimensionality are specified but do not match.
+
+ Returns:
+ Type[ArchiveSection]: The NumericalArray class with the specified shape and units.
+ '''
+ if shape is None:
+ if dimensionality is None:
+ dimensionality = 1
+ shape = ['*'] * dimensionality
+ else:
+ if dimensionality is not None and dimensionality != len(shape):
+ raise ValueError(f'Missmatch between dimensionality "{dimensionality}" and '
+ + f'shape "{shape}"')
+ array_shape = shape
+
+ class NumericalArray(ArchiveSection):
+ '''
+ A section for describing a numerical array with a normalizer for calculating
+ statistical descriptors.
+ '''
+ values = Quantity(
+ type=np.dtype(np.float64),
+ shape=array_shape,
+ description='The values of the numerical array',
+ unit=unit)
+
+ dimensionality = Quantity(
+ type=np.dtype(np.int8),
+ description='The dimensionality of the numerical array')
+
+ shape = Quantity(
+ type=np.dtype(np.int64),
+ shape=[len(array_shape)],
+ description='The shape of the numerical array')
+
+ mean = Quantity(
+ type=np.dtype(np.float64),
+ description='The mean value of the numerical array',
+ unit=unit)
+
+ minimum = Quantity(
+ type=np.dtype(np.float64),
+ description='The minimum value of the numerical array',
+ unit=unit)
+
+ maximum = Quantity(
+ type=np.dtype(np.float64),
+ description='The maximum value of the numerical array',
+ unit=unit)
+
+ standard_deviation = Quantity(
+ type=np.dtype(np.float64),
+ description='The standard deviation of the values in the numerical array',
+ unit=unit)
+
+ first_ventile = Quantity(
+ type=np.dtype(np.float64),
+ description='The first ventile of the values in the numerical array',
+ unit=unit)
+
+ first_quartile = Quantity(
+ type=np.dtype(np.float64),
+ description='The first quartile of the values in the numerical array',
+ unit=unit)
+
+ median = Quantity(
+ type=np.dtype(np.float64),
+ description='The median of the values in the numerical array',
+ unit=unit)
+
+ third_quartile = Quantity(
+ type=np.dtype(np.float64),
+ description='The third quartile of the values in the numerical array',
+ unit=unit)
+
+ nineteenth_ventile = Quantity(
+ type=np.dtype(np.float64),
+ description='The nineteenth ventile of the values in the numerical array',
+ unit=unit)
+
+ def normalize(self, archive, logger: Any) -> None:
+ '''The normalizer for the `NumericalArray` class. If values exist and the
+ statistical descriptors are not filled the normalizer will fill the
+ descriptors using NumPy properties, methods and functions.
+
+ Args:
+ archive (EntryArchive): The archive containing the section that is being
+ normalized.
+ logger (Any): A structlog logger.
+ '''
+ super(NumericalArray, self).normalize(archive, logger)
+
+ if self.values is not None:
+ if self.dimensionality is None:
+ self.dimensionality = self.values.ndim
+ if self.shape is None:
+ self.shape = self.values.shape
+ if self.mean is None:
+ self.mean = self.values.mean()
+ if self.minimum is None:
+ self.minimum = self.values.min()
+ if self.maximum is None:
+ self.maximum = self.values.max()
+ if self.standard_deviation is None:
+ self.standard_deviation = self.values.std()
+ if self.first_ventile is None:
+ self.first_ventile = np.quantile(self.values, 0.05)
+ if self.first_quartile is None:
+ self.first_quartile = np.quantile(self.values, 0.25)
+ if self.median is None:
+ self.median = np.quantile(self.values, 0.5)
+ if self.third_quartile is None:
+ self.third_quartile = np.quantile(self.values, 0.75)
+ if self.nineteenth_ventile is None:
+ self.nineteenth_ventile = np.quantile(self.values, 0.95)
+
+ return NumericalArray
+
+
class User(MSection):
user = Quantity(
type=author_reference,
--
GitLab