Skip to main content

Auto-Run Python Code at Startup

In previous chapters, we learned many Python embedded programming examples. Some of you may wonder: can we place already-written Python code on the Walnut Pi and have it execute automatically at startup? The answer is yes, and it's a very practical feature. In this section, we will explain it in detail.

The Walnut Pi system comes with an auto-run script feature that supports both desktop and headless (non-desktop) systems. No configuration is needed to execute specified script files at startup. For configuration instructions, please refer to: Auto-Run Scripts at Startup

Desktop System Method​

First, let's write a test Python code that uses the onboard LED blinking effect as a function demonstration, along with printing "Hello WalnutPi!" as a notification message. The code is as follows (available in the Walnut Pi development board resource package -- Example Programs -- Other Tips):

'''
Experiment Name: LED Blink
Experiment Platform: Walnut Pi
'''

# Import related modules
import board,time
from digitalio import DigitalInOut, Direction

# Build LED object and initialize
led = DigitalInOut(board.LED) # Define pin number
led.direction = Direction.OUTPUT  # IO as output

while True:

    led.value = 1 # Output high level, light up onboard blue LED
    
    time.sleep(0.5)
    
    led.value = 0 # Output low level, turn off onboard blue LED
    
    time.sleep(0.5)
    
    print("Hello WalnutPi!") # Output print message

Save the above code as led_blink.py and place it in the Walnut Pi /home/pi directory for testing. You can transfer it using Thonny or copy it directly to that directory using a USB drive.

auto_run1

Then, on the Walnut Pi desktop, click Start Menu -- Settings -- Session and Startup.

auto_run2

After launching, click Application Autostart, then click + Add at the bottom left. In the pop-up dialog, configure the following:

  • Name: Define the application name. We'll use START here; you can modify it as desired.
  • Description: Description of this application; can be left blank.
  • Command: The command. Here we use the following Python terminal command to launch the Python script. Be sure to use Python3 with proper capitalization, because this program has been granted elevated privileges on the Walnut Pi, with administrator rights to run Python scripts, avoiding permission-related errors.
Python3 /home/pi/led_blink.py
  • Trigger: Select "on login".

auto_run3

After confirmation, you can see the START program appear in the list.

auto_run4

Once created, you can find this startup file in the /home/pi/.config/autostart directory. Right-click and select Edit Launcher for more editing options.

auto_run5

As shown below, you can check the Run in terminal option, so that when the application opens, a new terminal will pop up on the desktop, making it easier to view debug information. If unchecked, it will run in the background.

auto_run6

After saving, you can double-click the application to test. You will see a new terminal pop up on the desktop, printing the output from led_blink.py.

auto_run7

The onboard LED blinks periodically!

auto_run8

Execute the reset command, and you will see the Walnut Pi development board pop up a terminal printing "Hello WalnutPi!" and the LED blinking, indicating that the Python code auto-start is working properly.

sudo reboot

Headless (Non-Desktop) System Method​

The previous section covered the method for the Walnut Pi desktop system. Here we'll explain the method for non-desktop versions, using the startup service method. This method also works for systems with a desktop!

As before, let's first write a test Python code using the onboard LED blinking effect for demonstration, along with printing "Hello WalnutPi!" as a notification message. The code is as follows (available in the Walnut Pi development board resource package -- Example Programs -- Other Tips):

'''
Experiment Name: LED Blink
Experiment Platform: Walnut Pi
'''

# Import related modules
import board,time
from digitalio import DigitalInOut, Direction

# Build LED object and initialize
led = DigitalInOut(board.LED) # Define pin number
led.direction = Direction.OUTPUT  # IO as output

while True:

    led.value = 1 # Output high level, light up onboard blue LED
    
    time.sleep(0.5)
    
    led.value = 0 # Output low level, turn off onboard blue LED
    
    time.sleep(0.5)
    
    print("Hello WalnutPi!") # Output print message

Save the above code as led_blink.py and place it in the Walnut Pi /home/pi directory for testing. You can transfer it using Thonny or copy it directly to that directory using a USB drive.

auto_run1

Create a start.service file in the /lib/systemd/system directory. (This file is available in the Walnut Pi development board resource package -- Example Programs -- Other Tips -- Non-Desktop System Method directory.)

sudo nano /lib/systemd/system/start.service

Modify the content as follows:

[Unit]
Description=Expand partition size

[Service]
Type=oneshot
ExecStart=Python3 /home/pi/led_blink.py
RemainAfterExit=yes
StandardOutput=null

[Install]
WantedBy=multi-user.target

After saving, grant maximum permissions to the file:

sudo chmod 777 /lib/systemd/system/start.service

auto_run10

Enable the service:

 sudo systemctl enable start.service

auto_run11

Execute the reset command. You will see the Walnut Pi development board's LED blinking after restart, indicating that the Python code auto-start is working properly. Since this method runs in the background, the "Hello WalnutPi!" print message won't be visible!

sudo reboot

auto_run8


tip

Both methods above work by launching a Python script via commands. Users can modify the command to run specific Linux commands or software on power-up.