Pass Duration Data Between MATLAB Compiler SDK and Python

Supported platforms: Windows®, Linux®, Mac

This example shows how to create a Python® application that uses MATLAB® Compiler SDK™ to pass duration data between Python and MATLAB. The Python script calls two packaged MATLAB functions that use duration data as input or output.

When you pass a MATLAB duration scalar to Python, MATLAB Compiler SDK converts the data into a datetime.timedelta object. When you pass a MATLAB duration array to Python, MATLAB Compiler SDK converts the data into a matlab.object array if NumPy is not available in the Python environment, or into a numpy.timedelta64 array if NumPy is available in the Python environment. When you pass a Python datetime.timedelta or numpy.timedelta64 to MATLAB, MATLAB Compiler SDK converts the data into a duration. For a complete list of data types and their conversions, see Pass Data Between MATLAB and Python.

Prerequisites

Verify that you have a version of Python installed that is compatible with MATLAB Compiler SDK. For details, see MATLAB Compiler SDK Python Target Requirements.

The target machine requires MATLAB Runtime to run the deployed MATLAB code. You can include MATLAB Runtime in an installer for your Python package or install it separately. For details, see Download and Install MATLAB Runtime.

The NumPy package is not required for MATLAB Compiler SDK in general, but allows you to convert MATLAB Runtime duration arrays to Python numpy.timedelta64 arrays.

Create MATLAB Function with duration Input

This MATLAB function takes a duration as input. It calculates the longest duration in a duration array. Save this function as getLongestDuration.m.

function longestDuration = getLongestDuration(durArray)
%   Given an array of durations, return the longest duration
    longestDuration = max(durArray);
end

Create MATLAB Function with duration Output

This MATLAB function returns a duration. This function finds the duration between two datetimes. Save this function as durAtoB.m.

function durBetween = durAtoB(dtA, dtB)
%  Given two datetimes, find the duration
%  between them.

%  Since between() returns a calendarDuration,
%  convert it to a duration
    calDurBetween = between(dtA, dtB);
    durBetween = time(calDurBetween);
end

Create Python Package from Function

Build the Python package with the Python Package Compiler app or the compiler.build.pythonPackage function.

For example, if you are using compiler.build.pythonPackage, start by getting a list of files with a .m extension in the current directory.

functionfiles = dir('*.m');
Next, save the filenames in a cell array.
functionfiles = {functionfiles.name};
Then, compile the Python package.
buildResults = compiler.build.pythonPackage(functionfiles,'PackageName','durationLib');

For details, see the instructions in Generate Python Package and Build Python Application.

Write Python Application Code

Write code in Python for an application named testDurationFunctions.py. This sample application contains driver code to test the packaged MATLAB functions.

#!/usr/bin/env python

import durationLib

# Import the matlab module only after you have imported
# MATLAB Compiler SDK generated Python modules.
import matlab
import datetime as dt
import numpy as np

try:
    # Initialize the Python module created with MATLAB Compiler SDK.
    my_durationLib = durationLib.initialize()
except Exception as e:
    print("Error initializing durationLib package\\n:{}".format(e))
    exit(1)

try:
    # Create the duration array.
    string_list = [
        dt.datetime.strptime(i, "%H:%M:%S")
        for i in [       
                         "02:23:41",
                         "17:46:39",
                         "03:47:10",
                         "06:12:08",
                         "12:29:30",
                         "03:27:57",
                         "02:39:21",
                         "21:42:35",
                         "15:45:14",
                         "22:17:45"
            ]
        ]

    timedelta_arr = np.array([np.timedelta64(t.hour, 'h')+
                               np.timedelta64(t.minute, 'm')+ 
                               np.timedelta64(t.second, 's')
                      for t in string_list])
    # Call the packaged functions.
    yOut = my_durationLib.getLongestDuration(timedelta_arr)
    print("The longest timedelta in the list is: " + str(yOut))

    yOut = my_durationLib.durAtoB(dt.datetime(2024, 10, 31, 18, 0), 
                                  dt.datetime(2024, 10, 31, 15, 0))
    print("The timedelta between the two datetimes is: " + str(yOut))
    
except Exception as e:
    print("Error occurred during program execution\\n:{}".format(e))

my_durationLib.terminate()

The application performs these steps:

  • Imports the durationLib, matlab, datetime and numpy packages.

  • Instantiates the durationLib instance as my_durationLib.

  • Creates a list of durations called duration_list.

  • Converts the list to a timedelta64 NumPy array called timedelta_arr.

  • Calls the packaged functions longestDuration and durAtoB.

  • Displays the result.

  • Uses a try block to catch and handle exceptions.

Install and Run Python Package

On the target machine, install the generated Python package. For more details, see Install and Import MATLAB Compiler SDK Python Packages.

In the system command prompt, navigate to the folder containing the generated files and install the Python package.

python -m pip install .

Run the application.

python testDurationFunctions.py

The application generates this output.

The longest timedelta in the list is: 22:17:45
The timedelta between the two datetimes is: -1 day, 21:00:00

See Also

Topics