Bulk-Update Using a Raspberry Pi Board

This example uses:

    Open Live Script

    This example shows how to use a Raspberry Pi™ board that runs Python® 2.7 that is connected to a Wi-Fi® network to collect data. You can continuously collect CPU temperature and percentage of CPU utilization over 15 seconds and bulk-update a ThingSpeak™ channel every 2 minutes. This example uses the Bulk-Write JSON Data API to collect data as a batch and send it to ThingSpeak channels. This bulk-update reduces the power usage of your devices. Since Raspberry Pi board does not come with a real-time clock, you can use the relative time stamp for bulk-update messages.

    Setup

    1) Create a channel as shown in Collect Data in a New Channel.

    2) Import the necessary libraries for the script.import urllib2 as ul

    import json
import time
import os
import psutil

    3) Define global variables that track the last connection time and last update time. Also, define time intervals to update the data, and post the data to ThingSpeak.

    lastConnectionTime = time.time() # Track the last connection time
lastUpdateTime = time.time() # Track the last update time
postingInterval = 120 # Post data once every 2 minutes
updateInterval = 15 # Update once every 15 seconds

    4) Define your ThingSpeak channel settings such as write API key and channel ID along with ThingSpeak server settings.

    writeAPIkey = "YOUR-CHANNEL-WRITEAPIKEY" # Replace YOUR-CHANNEL-WRITEAPIKEY with your channel write API key
channelID = "YOUR-CHANNELID" # Replace YOUR-CHANNELID with your channel ID
url = "https://api.thingspeak.com/channels/"+channelID+"/bulk_update.json" # ThingSpeak server settings
messageBuffer = []

    5) Define the function httpRequest to send data to ThingSpeak and to print the response code from the server. A response code 202 indicates that the server has accepted the request and will process it.

    def httpRequest():
	'''Function to send the POST request to 
	ThingSpeak channel for bulk update.'''

	global messageBuffer
	data = json.dumps({'write_api_key':writeAPIkey,'updates':messageBuffer}) # Format the json data buffer
	req = ul.Request(url = url)
	requestHeaders = {"User-Agent":"mw.doc.bulk-update (Raspberry Pi)","Content-Type":"application/json","Content-Length":str(len(data))}
	for key,val in requestHeaders.iteritems(): # Set the headers
		req.add_header(key,val)
	req.add_data(data) # Add the data to the request
    # Make the request to ThingSpeak
	try:
		response = ul.urlopen(req) # Make the request
		print response.getcode() # A 202 indicates that the server has accepted the request
	except ul.HTTPError as e:
		print e.code # Print the error code
	messageBuffer = [] # Reinitialize the message buffer
	global lastConnectionTime
	lastConnectionTime = time.time() # Update the connection time

    6) Define the function getData that returns the CPU temperature in Celsius along with the CPU utilization as a percentage.def getData():

    '''Function that returns the CPU temperature and percentage of CPU utilization'''
	cmd = '/opt/vc/bin/vcgencmd measure_temp'
	process = os.popen(cmd).readline().strip()
	cpuTemp = process.split('=')[1].split("'")[0]
	cpuUsage = psutil.cpu_percent(interval=2)
	return cpuTemp,cpuUsage

    7) Define the function updatesJson to continuously update the message buffer every 15 seconds.

    def updatesJson():
	'''Function to update the message buffer
	every 15 seconds with data. And then call the httpRequest 
	function every 2 minutes. This examples uses the relative timestamp as it uses the "delta_t" parameter. 
	If your device has a real-time clock, you can also provide the absolute timestamp using the "created_at" parameter.
	'''

	global lastUpdateTime
	message = {}
	message['delta_t'] = int(round(time.time() - lastUpdateTime))
	Temp,Usage = getData()
	message['field1'] = Temp
	message['field2'] = Usage
	global messageBuffer
	messageBuffer.append(message)
    # If posting interval time has crossed 2 minutes update the ThingSpeak channel with your data
	if time.time() - lastConnectionTime >= postingInterval:
		httpRequest()
	lastUpdateTime = time.time()

    8) Run an infinite loop to continuously call the function updatesJson once every 15 seconds.

    if __name__ == "__main__":  # To ensure that this is run directly and does not run when imported 
	while 1:
		# If update interval time has crossed 15 seconds update the message buffer with data
	    if time.time() - lastUpdateTime >= updateInterval:
	        updatesJson()

    Related Examples

    More About

    ThingSpeak Documentation
    Support
    Introducing Deep Learning with MATLAB

    Introducing Deep Learning with MATLAB

    Download ebook