Granta MI Scripting Toolkit 4.2

Edit Tabular Data

Last update: 16.07.2025

Edit and compare tabular data using the Python numpy library.

In this example, there is some existing tensile data for a material (AMS 6520, Plate, 1000°F). Some of that data is of particular interest and we want to separate it from the old data and put it in a record of its own.

Connect to MI

Specify a database and table.

from datetime import datetime
import numpy as np
from GRANTA_MIScriptingToolkit import granta as mpy

mi = mpy.connect("http://my.server.name/mi_servicelayer", autologon=True)

db = mi.get_db(db_key="MI_Training")
table = db.get_table("Tensile Statistical Data")

Locate existing data

Find the record and attribute that contains the original data.

record = table.search_for_records_by_name("AMS 6520, Plate, 1000°F")[0]
samples = record.attributes["Tensile test data used in this rollup"]

Select some samples of interest.

focus_samples = ["MTS-615731", "MTS-615741", "MTS-615771"]

Get their Record objects from the tabular attribute containing the tensile test data.

focus_recs = [samples.linked_records[s][0] for s in focus_samples]

Get their row indices in the tabular data structure.

focus_rows = [samples["Linking Value (Specimen ID)"].index(samp) for samp in focus_samples]

Extract and analyse the data

Extract the sample data, and establish how different these particular samples are from the remainder of the set.

focus_youngs = [samples["Young's Modulus (11-axis)", i] for i in focus_rows]
rmain_youngs = [samples["Young's Modulus (11-axis)", i] for i in range(6) if i not in focus_rows]

rmain_mean = np.mean(rmain_youngs)
focus_mean = np.mean(focus_youngs)
rmain_mean, focus_mean

Previous cell output:

(np.float64(142.74677022298178), np.float64(146.3169097900391))

percentage_diff = 100*abs(rmain_mean-focus_mean)/(0.5*(rmain_mean+focus_mean))
print(f"Percentage difference between the two means is: {percentage_diff} %")

Previous cell output:

Percentage difference between the two means is: 2.470140535743884 %

Create a new record and tabular attribute

Create a new record for your data of interest.

now = datetime.now().strftime("%c")
recordName = f"Scripting Toolkit Example 9:{now}"
new_rec = table.create_record(recordName, parent=record.parent, subsets={"Statistical Test Data"})

Select the tabular attribute to write to.

new_samples = new_rec.attributes["Tensile test data used in this rollup"]
new_samples

Previous cell output:

<tabularvalue name: tensile test data used in this rollup>

Link the samples to the new tabular datum.

for focus_sample in focus_samples:
    new_samples.add_row(linking_value=focus_sample)
new_samples

Previous cell output:

<tabularvalue name: tensile test data used in this shape: x>

Set the new tabular attribute to update, and write the new record to MI.

new_rec.set_attributes([new_samples])
new_rec = mi.update([new_rec])[0]

The original record is linked to the records you just linked to the tabular data. Link the new record to the original, too.

(Records must exist on the server to be linked together, which means the record must be pushed to the server before it can be linked to other records.)

new_rec.set_links("Tensile Test Data", set(focus_recs))
new_rec.links

Previous cell output:

{'Tensile Test Data': {<record long name: mts-615731>,
  <record long name: mts-615741>,
  <record long name: mts-615771>}}

Like changes to data, changes to links also need to be pushed to the server. Unlike attributes, links do not need to be flagged for update (there is no equivalent to set_attributes() for links).

mi.update_links([new_rec])

Previous cell output:

[<record long name: scripting toolkit example may>]

Connect with Ansys