Does the dataset parameter of the evaluate function have to come from a solution?

[bearmeetu]

I simulated the electric field strength of the transmission line using COMSOL and wanted to use the evaluate function to calculate the electric field strength value at a three-dimensional intersection point (three-dimensional intersection point 2), but the code reported an error: Dataset "three-dimensional intersection point 2" does not refer to a solution. Does the dataset parameter of the evaluate function have to come from a solution? Here is my code and mph model file.

'import mph
def export_electric_field(model_path, output_file):
try:
# 启动 COMSOL 客户端
client = mph.start(cores=1)
print("COMSOL 客户端启动成功。")

    # 加载模型
    model = client.load(model_path)
    print(f"模型加载成功：{model_path}")

    parameters = model.parameters()
    print("参数：", parameters)

    materials = model.materials()
    print("材料：", materials)

    physics = model.physics()
    print("物理：", physics)

    studies = model.studies()
    print("研究：", studies)

    # 求解模型
    model.solve('研究 1')
    print("仿真运行完成")

    mph.tree(model)

    datasets = model.datasets()
    print("数据集：", datasets)

    model.export('数据 1', output_file)
    print("数据已导出：", output_file)

    #res = model.evaluate('esbe.normE', dataset='研究 1//解 1')
    res = model.evaluate('esbe.normE', unit='V/m', dataset='三维截点 2')
    print("电场强度结果：", res)

    client.remove(model)

except Exception as e:
    print(f"运行仿真时出错：{e}")

if name == "main":
model_file_path = "/home/icebear/comsol/mphs/powerline.mph"
output_file = "electric_field_data.txt"
export_electric_field(model_file_path, output_file)'

comsol mph model file: powerline.mph

[john-hen]

Yes.

Other dataset types are not supported. The idea was discussed in issue #26, but has since been abandoned.

You can define an appropriate integration coupling operator in the model, one that operates on your (non-solution) dataset. The result is then available as a global evaluation in the default dataset and can thus be retrieved with model.evaluate().