REQUEST: Query model about rfem_python_client

[OndraMichal]

Hi @ZTThomasOberthanner,
thank you for thr question. To query the model you start with initializing the connection. This can be done via Model(False, "model_name"). Then use any getter you need to retrieve the data. Our unit tests (RFEM_Python_Client/UnitTests/) are good examples of setting values and getting them back for checking correctness. But in the tests we use currently available (active) model, that is why you see just Model() there.
If you want to see all available model functions and types (prefix ns0:), just initialize class Model and then print(Model.clientModel). This is the easiest way SUDS module shows everything it recognizes.
I'll add this process to our Wiki.

[ZTThomasOberthanner]

Thanks for the quick reply, i did had a look at test_Member_test.py but I could not find any tests for correctness there.
I am now able to get node/member information from the model. But I cannot figure out how to get the internal forces for a member for a certain loadcase/result combination. Would by happy for another hint.

[DavidNaizheZhou]

is it possible to query a list/table of all nodes, members, ...?

suggestion:
get_node() - function with a default None

get_node(xs=None) -> provides the nodes table as (DataFrame or a Dict)

[ZTThomasOberthanner]

this is my implementation for fetching all nodes

    def get_nodes(self):
        numbers=self.service.get_all_object_numbers(type='E_OBJECT_TYPE_NODE')
        return [self.service.get_node(n) for n in numbers]

you cannot copy paste that, because this is part of my own python client, but i think you get the idea...

[DavidNaizheZhou]

got it...thank you!!

here is my minimal working example:

from RFEM.initModel import Model, ConvertStrToListOfInt
from RFEM.BasicObjects.node import Node

Model(new_model=True, model_name="extract")

# model modification
Model.clientModel.service.begin_modification()
Node(1, 0.0, 0.0, 0.0)
Node(2, 1.0, 0.0, 0.0)
Node(3, 2.0, 0.0, 0.0)
Model.clientModel.service.finish_modification()

# extract all nodes
numbers = ConvertStrToListOfInt(
    Model.clientModel.service.get_all_object_numbers(type="E_OBJECT_TYPE_NODE")
)
node_list = [Model.clientModel.service.get_node(n) for n in numbers]

[dogukankaratas]

Hey, thanks for bringing this up and ideas. I would like to jump back the question and drop here a suggestion. Here is a function to create DataFrame tables for Node objects.

def getNodeTable():
        import pandas as pd
        nodeTable = pd.DataFrame(columns=["Node No.", "X", "Y", "Z"])
        modelStatus = GetModelInfo()
        for i in range(1, modelStatus[0] + 1):
            nodeTable.loc[i] = [i, Model.clientModel.service.get_node(i)['coordinate_1'],
                                   Model.clientModel.service.get_node(i)['coordinate_2'],
                                   Model.clientModel.service.get_node(i)['coordinate_3']]

        return nodeTable

The same logic will work for another objects (lines, members, surfaces etc.).

Maybe this functions can be added to our static methods. Please let me know about your thoughts ✌️

[DavidNaizheZhou]

here is my suggestion for fetching tables into a pandas dataframe:

import pandas as pd
import numpy as np

from RFEM.initModel import Model, ConvertStrToListOfInt
from RFEM.BasicObjects.node import Node
from RFEM.BasicObjects.material import Material
from RFEM.BasicObjects.section import Section


def load_object(clientModel, type="NODE"):
    numbers = ConvertStrToListOfInt(
        clientModel.service.get_all_object_numbers(
            type="E_OBJECT_TYPE_{}".format(type.upper())
        )
    )
    if type.upper() == "NODE":
        return [Model.clientModel.service.get_node(int(n)) for n in numbers]
    else:
        # todo object types
        pass


def load_dataframe(clientModel, type="NODE"):
    nodes_list = load_object(clientModel, type)
    len_numbers = len(nodes_list)
    keys = dict(nodes_list[0]).keys()
    dataframe = pd.DataFrame(columns=keys, index=range(len_numbers), dtype=object)
    for i, node in enumerate(nodes_list):
        node_dict = dict(node)
        values = node_dict.values()
        dataframe.iloc[i] = np.array([value for value in values], dtype=object)
    return dataframe


Model(new_model=True, model_name="extract_dataframe")

# model modification
Model.clientModel.service.begin_modification()
Node(1, 0.0, 0.0, 0.0)
Node(2, 1.0, 0.0, 0.0)
Node(3, 2.0, 0.0, 0.0)
Model.clientModel.service.finish_modification()

# load dataframe
df = load_dataframe(Model.clientModel, type="NODE")
print(df)

[ZTThomasOberthanner]

just a reminder:
clientModel.service.get_all_object_numbers already returns a list of int - so no need for ConvertStrToListOfInt...

numbers = ConvertStrToListOfInt(
        clientModel.service.get_all_object_numbers(
            type="E_OBJECT_TYPE_{}".format(type.upper())
        )
    )

[dogukankaratas]

Since everyone has a solution, this issue will be closed 😃 Thanks for contribution.