Fields Container and Fields¶
Where DPF uses Operators to load operate and operate
on the data, DPF uses Field and FieldsContainer to store
the data and return it to the user. In other words, if Operators
was a verb, the DPF Field would be the noun. Operators acts
on the data, and Field and FieldsContainer hold the data.
Obtaining Fields and FieldsContainer¶
The outputs from Operators can be either a
ansys.dpf.core.field.Field or
ansys.dpf.core.fields_container.FieldsContainer. A Fields
Container is the DPF equivalent of a list of Field, and is
used to hold a vector of fields within DPF. In the following the
example, the FieldsContainer is returned from the elastic_strain operator:
from ansys.dpf import core as dpf
from ansys.dpf.core import examples
model = dpf.Model(examples.msup_transient)
epel = model.results.elastic_strain.on_all_time_freqs
fields = epel.eval()
print(fields)
Out:
DPF elastic_strain(s)Fields Container
with 20 field(s)
defined on labels: time
with:
- field 0 {time: 1} with ElementalNodal location, 6 components and 40 entities.
- field 1 {time: 2} with ElementalNodal location, 6 components and 40 entities.
- field 2 {time: 3} with ElementalNodal location, 6 components and 40 entities.
- field 3 {time: 4} with ElementalNodal location, 6 components and 40 entities.
- field 4 {time: 5} with ElementalNodal location, 6 components and 40 entities.
- field 5 {time: 6} with ElementalNodal location, 6 components and 40 entities.
- field 6 {time: 7} with ElementalNodal location, 6 components and 40 entities.
- field 7 {time: 8} with ElementalNodal location, 6 components and 40 entities.
- field 8 {time: 9} with ElementalNodal location, 6 components and 40 entities.
- field 9 {time: 10} with ElementalNodal location, 6 components and 40 entities.
- field 10 {time: 11} with ElementalNodal location, 6 components and 40 entities.
- field 11 {time: 12} with ElementalNodal location, 6 components and 40 entities.
- field 12 {time: 13} with ElementalNodal location, 6 components and 40 entities.
- field 13 {time: 14} with ElementalNodal location, 6 components and 40 entities.
- field 14 {time: 15} with ElementalNodal location, 6 components and 40 entities.
- field 15 {time: 16} with ElementalNodal location, 6 components and 40 entities.
- field 16 {time: 17} with ElementalNodal location, 6 components and 40 entities.
- field 17 {time: 18} with ElementalNodal location, 6 components and 40 entities.
- field 18 {time: 19} with ElementalNodal location, 6 components and 40 entities.
- field 19 {time: 20} with ElementalNodal location, 6 components and 40 entities.
Accessing Fields within a Fields Container¶
Since this result contains a transient result, the
FieldsContainer has one Field by time set. Access the fields
from the FieldsContainer using the following methods:
len(fields)
Out:
20
Return a field based on its index
field_first_time = fields[0]
field_last_time = fields[19]
Return a field based on its time set id:
field = fields.get_field_by_time_id(1)
Alternatively, use get_field with the identifier of
the requested field. This API allows to access fields for more complex
requests. To use get_field in the same context:
field = fields.get_field({'time': 1})
print(field)
Out:
DPF elastic_strain_0.01s Field
Location: ElementalNodal
Unit:
40 entities
Data:6 components and 320 elementary data
Or in a more complex context, where the get_field method has a real use compared
to the get_field_by_time_id method:
model = dpf.Model(examples.download_all_kinds_of_complexity())
epel = model.results.elastic_strain.on_all_time_freqs.split_by_shape
fields = epel.eval()
field = fields.get_field({'time': 1, 'elshape':0})
print(field)
field = fields.get_field({'time': 1, 'elshape':1})
print(field)
Out:
DPF elastic_strain_1.s_elshape:0 Field
Location: ElementalNodal
Unit:
203 entities
Data:6 components and 2436 elementary data
DPF elastic_strain_1.s_elshape:1 Field
Location: ElementalNodal
Unit:
9052 entities
Data:6 components and 37580 elementary data
Reference the available time frequency support to determine which
time_complex_ids are available to in the FieldsContainer
model = dpf.Model(examples.msup_transient)
epel = model.results.elastic_strain.on_all_time_freqs
fields = epel.eval()
print(fields.time_freq_support)
Out:
DPF Time/Freq Support:
Number of sets: 20
Cumulative Time (s) LoadStep Substep
1 0.010000 1 1
2 0.020000 1 2
3 0.030000 1 3
4 0.040000 1 4
5 0.050000 1 5
6 0.060000 1 6
7 0.070000 1 7
8 0.080000 1 8
9 0.090000 1 9
10 0.100000 1 10
11 0.110000 1 11
12 0.120000 1 12
13 0.130000 1 13
14 0.140000 1 14
15 0.150000 1 15
16 0.160000 1 16
17 0.170000 1 17
18 0.180000 1 18
19 0.190000 1 19
20 0.200000 1 20
Note that the time set ids used are 1 based. When indexing from pythonic indexing via
fields[0], you can use zero based indexing. When requesting the
results via get_fields, the request is based on the time scoping set ids.
Field¶
The ansys.dpf.core.field.Field is the fundamental unit of data within DPF.
It contains the actual data and its metadata:
results data are defined by values associated to entities (scoping), and these entities are a subset of a model (support).
In DPF, field data is always associated to its scoping and support, making the field a self-describing piece of data.
A field is also defined by its dimensionnality, unit, location…
field = fields[0]
print(field)
Out:
DPF elastic_strain_0.01s Field
Location: ElementalNodal
Unit:
40 entities
Data:6 components and 320 elementary data
Note that by printing the field you can get an overview of the field’s metadata. The next section provides an overview of the metadata associated with the field itself.
Field Metadata¶
The field contains the metadata associated with the result it is
associated with, including the location (either Elemental, Nodal,
ElementalNodal, …) and IDs associated with that location. To access the
scoping of the field, use the scoping attribute:
>>> print(field.scoping)
>>> print('field.scoping.ids:', field.scoping.ids)
>>> print('field.location:', field.location)
Out:
DPF Scoping:
with Elemental location and 40 entities
field.scoping.ids: [21,
22,
23,
24,
25,
26,
...
]
field.location:'ElementalNodal'
Note that Elemental denotes one value (multiplied by the number of
components) of data per element, Nodal is per node, and
ElementalNodal is one value per node per element. For example,
strain is a ElementalNodal value as the strain is evaluated at
each node for each element.
The field also contains additional metadata such as the shape of
the data stored, the location of the field, number of components, and
the units of the data,
>>> stress = model.results.stress
>>> field = stress.eval()[0]
Units of the field describing volume
>>> field.unit
Elemental, elemental nodal, or nodal element "location" of the field
>>> field.location
Number of components associated with the field. It's expected to
be have a single dimension since there can only be one volume per
element.
>>> field.component_count
Out:
'Pa'
'ElementalNodal'
6
Field Data¶
Accessing Field Data¶
When DPF-Core returns a ansys.dpf.core.field.Field, what
Python actually has is a client side representation of the Field, but
not the entirety of the field itself. This means that all the data of
the field is stored within the DPF service. This is important because
when building your post-processing workflows, keep in mind that the
most efficient way of interacting with the result data is to minimize
the exchange of data between Python and DPF, either by using operators
or by accessing only the data that is needed.
Should you need to access the entire array of data, you can request
the data be returned as a numpy array with:
>>> array = field.data
>>> array
array([[ 4.01372930e+04, 3.85071930e+02, -1.40019130e+07,
7.48472351e+02, -2.60259531e+04, -2.62856938e+05],
[-1.19228638e+03, -6.18210815e+02, -1.39912700e+07,
2.61468994e+03, -1.31871719e+05, -2.59527125e+05],
[ 9.02558960e+02, 5.63793152e+02, -1.17102740e+07,
-8.99381836e+02, -1.21302727e+05, -2.45666328e+05],
...,
[-3.99694531e+04, 1.44622528e+02, 9.62343100e+06,
-7.09812073e+02, -2.26106621e+04, -2.23155891e+05],
[-4.31104401e+02, -2.67612732e+02, 9.60954800e+06,
1.93208755e+02, -1.11580734e+05, -2.24406062e+05],
[ 5.56899536e+02, 3.88515320e+02, 1.17119880e+07,
-1.68983887e+03, -1.21768023e+05, -2.41346125e+05]])
This array has 6 components by elementary data (symmetrical tensor XX,YY,ZZ,XY,YZ,XZ)
Note that this array is a genuine, local, numpy array
>>> type(array)
numpy.ndarray
Should you need to request data individual node or element, you can
request it using either get_entity_data or
get_entity_data_by_id.
Get the data from the first element in the field.
>>> field.get_entity_data(0)
Get the data for the element with the ID 10
>>> field.get_entity_data_by_id(10)
array([[ 4.99232031e+04, 1.93570602e+02, -3.08514075e+06,
-5.48255615e+02, -1.37476562e+04, 1.34827719e+05],
[ 5.23090469e+04, -1.87847885e+02, -1.98004588e+06,
-1.12942969e+03, -1.11147285e+04, 1.09223398e+05],
[-4.90510511e+00, -1.16425255e+02, -1.96296662e+06,
-5.48878540e+02, -5.48524844e+04, 1.09255164e+05],
[ 2.63994884e+01, 1.50431015e+02, -3.06906050e+06,
-1.17046680e+03, -6.76924219e+04, 1.34773391e+05],
[-4.99232031e+04, -1.93571167e+02, 3.08514075e+06,
-5.48256836e+02, -1.37476562e+04, -1.34827719e+05],
[-5.23090469e+04, 1.87848083e+02, 1.98004588e+06,
-1.12943201e+03, -1.11147295e+04, -1.09223398e+05],
[ 4.90471840e+00, 1.16423714e+02, 1.96296662e+06,
-5.48877380e+02, -5.48524844e+04, -1.09255164e+05],
[-2.63994102e+01, -1.50429443e+02, 3.06906050e+06,
-1.17046619e+03, -6.76924219e+04, -1.34773391e+05]])
Note that this would correspond to an index of 29 within the
field. Be aware that scoping IDs are not sequential. The index
of Element 29 in the field can be obtained by:
>>> field.scoping.ids.index(10)
29
Here the data of element of id 10 is made of 8 symmetrical tensor, indeed
the elastic strain has one tensor value by node by element (``ElementalNodal`` location)
For a displacement on node 3, we have :
>>> disp = model.results.displacement.eval()[0]
>>> disp.get_entity_data_by_id(3)
array([[8.06571808e-14, 4.03580652e-04, 2.61804706e-05]])
One 3D vector (X,Y,Z) displacement
These methods are acceptable when requesting data for a few elements or nodes, but should not be used when looping over the entire array. Field’s data can be recovered locally before sending a large number of requests for efficiency purpose:
with field.as_local_field() as f:
for i in range(1,100):
f.get_entity_data_by_id(i)
Operating on Field Data¶
Often times, it’s not necessary to directly act upon the data of an
array within Python. For example, if you want to know the maximum of
the data, you could potentially compute the maximum of the array from
numpy with array.max(), but that requires sending the entire
array to Python and then computing the maximum there. Rather than
copying the array over and then computing the maximum in Python, you
can instead compute the maximum directly from the field itself. The
min and max methods of Field use the 'min_max'
Operator was used to compute the maximum of the field while returning
the field.
Compute the maximum of the field within DPF and return the result
a numpy array
>>> max_field = field.max()
>>> max_field.data
array([0.12492393, 0.06738043, 0.05854268, 0.05807593, 0.08250141,
0.2068032 ])
Get the element or node ID of the maximum value.
>>> max_field.scoping.ids
[369, 1073, 1031, 1040, 2909, 2909]
Note that is a convenience method. For more advanced operator
chaining, please see the Operators section in
the user guide. Here is a quick example of connecting the
'elemental_mean' operator to the field data:
Compute the average of a field using the ‘elemental_mean’ operator:
from ansys.dpf.core import operators as ops
avg_op = ops.averaging.elemental_mean(field)
avg_field = avg_op.outputs.field()
print(avg_field.get_entity_data(0))
print(avg_field.location)
Out:
[[ 4.65393066e-04 -2.47955322e-05 0.00000000e+00 7.68026390e+02
-7.59655688e+04 0.00000000e+00]]
Elemental
API Reference¶
See the API reference at FieldsContainer and Field.