Live Plot of CV Data

[Steffen]

Hi

I was wondering if there is an easy way I can plot the measurement data of the Rodeostat live while running e.g. a CV measurement? It should be noted that I am not in any way experienced when it comes to Python programming and I am also only a beginner when it comes to Arduino programming.
I have managed to create a data collection system with live data plotting for our old manual potentiostat/galavanostat using Arduino and C# but I chose to order the Rodeostat instead of trying to develop a similar microcontroller based system myself.

Best Regards
Steffen Vindt

[Will Dickson]

@Steffen

Yes. The easiest way is probably to create a sub-class of the Potentiostat class with it own run_test method which creates the live plot instead of the progress bar. I give a simple example below.

from __future__ import print_function
import json
import threading
import Queue
import matplotlib.pyplot as plt
import potentiostat as ps


class PotentiostatLivePlot(ps.Potentiostat):
    """
    Custom version of potentiostat with displays a live plot when running a test.
    """

    def __init__(self, port, timeout=10.0, debug=False):
        super(PotentiostatLivePlot,self).__init__(port,timeout=timeout,debug=debug)
        self.data_queue = Queue.Queue()


    def get_default_volt_lims(self):
        volt_range = self.get_volt_range()
        volt_range_value = float(volt_range[:-1])
        volt_lims = -volt_range_value, volt_range_value
        return volt_lims


    def get_default_curr_lims(self):
        curr_range = self.get_curr_range()
        curr_range_value = float(curr_range[:-2])
        curr_lims = -curr_range_value, curr_range_value
        return curr_lims


    def get_curr_unit(self):
        curr_range = self.get_curr_range()
        curr_unit = curr_range[-2:]
        return curr_unit


    def receive_data(self,timeunit):
        while True:
            # Get json data from the device
            sample_json = self.readline()
            sample_json = sample_json.strip()
            try:
                sample_dict = json.loads(sample_json.decode())
            except ValueError:
                continue
            # Put new values in data queue
            if len(sample_dict) > 0:
                tval = sample_dict[ps.TimeKey]*ps.TimeUnitToScale[timeunit]
                volt = sample_dict[ps.VoltKey]
                curr = sample_dict[ps.CurrKey]
                self.data_queue.put({'tval': tval, 'volt': volt, 'curr': curr})
            else:
                self.data_queue.put({})



    def run_test(self, testname, param=None, filename=None, volt_lims=None, curr_lims=None, timeunit='s'):

        if timeunit not in ps.TimeUnitToScale:
            raise RuntimeError('uknown timeunit option {0}'.format(timeunit))

        if volt_lims is None:
            volt_lims = self.get_default_volt_lims()

        if curr_lims is None:
            curr_lims = self.get_default_curr_lims()

        curr_unit = self.get_curr_unit() 

        if param is not None:
            self.set_param(testname,param)

        if filename is not None:
            fid = open(filename,'w')

        t_done = self.get_test_done_time(testname, timeunit=timeunit)

        # Create figure and live plot
        fig = plt.figure()
        plt.ion()

        volt_axes = plt.subplot(211)
        volt_line, = plt.plot([0],[0],'b')
        plt.xlim(0,t_done)
        plt.ylim(*volt_lims)
        plt.grid('on')
        plt.ylabel('(V)')
        plt.title('Potentiostat Live Plot')

        curr_axes = plt.subplot(212,sharex=volt_axes)
        curr_line, = plt.plot([0],[0],'b')
        plt.xlim(0,t_done)
        plt.ylim(*curr_lims)
        plt.grid('on')
        plt.ylabel('({0})'.format(curr_unit))
        plt.xlabel('({0})'.format(timeunit))

        fig.canvas.flush_events()
        plt.pause(0.001)

        # Create lists for storing incoming data
        tval_list = [] 
        volt_list = [] 
        curr_list = []

        # Setup up working thread to collect data from device
        data_worker = threading.Thread(target=self.receive_data,args=(timeunit,))
        data_worker.daemon = True

        # Send command to run the test
        cmd_dict = {
                ps.CommandKey: ps.RunTestCmd, 
                ps.TestKey: testname
                }
        msg_dict = self.send_cmd(cmd_dict)
        self.test_running = True

        data_worker.start()
        done = False

        while not done:

            # Get available data from data queue and add to list
            while self.data_queue.qsize() > 0:
                data = self.data_queue.get(True)
                if not data:
                    done = True # Empty data means run is complete
                    break

                # Get data items and append to lists
                tval = data['tval']
                volt = data['volt']
                curr = data['curr']
                tval_list.append(tval)
                volt_list.append(volt)
                curr_list.append(curr)

                # Write data to file
                if filename is not None:
                    fid.write('{0:1.3f}, {1:1.4f}, {2:1.4f}\n'.format(tval,volt,curr))

                print('{0:1.3f}, {1:1.4f}, {2:1.4f}'.format(tval,volt,curr))

            # Update live plot
            volt_line.set_xdata(tval_list)
            volt_line.set_ydata(volt_list)
            curr_line.set_xdata(tval_list)
            curr_line.set_ydata(curr_list)
            fig.canvas.flush_events()
            plt.pause(0.001)

        self.test_running = False

        if filename is not None:
            fid.close()

        return tval_list, volt_list, curr_list 

 

Assuming you have saved this to a new module called potentiostat_live_plot.py then you could call it as follows.

 

from potentiostat_live_plot import PotentiostatLivePlot

dev = PotentiostatLivePlot(port='/dev/ttyACM0')

dev.set_curr_range('100uA')
dev.set_sample_period(10)

name = 'cyclic'
param = {
        'quietValue' : 0.0,
        'quietTime'  : 1000,
        'amplitude'  : 2.0,
        'offset'     : 0.0,
        'period'     : 1000,
        'numCycles'  : 15,
        'shift'      : 0.0,
        }

t, volt, curr = dev.run_test(name,param,filename='data.txt', curr_lims=(-50,50))

raw_input('Press enter to quit')

 

The live plot should look something like the show image below.

 

 

In this example there are two plots which are updated in realtime - one showing the output voltage vs time and the other showing the current vs time.

[Steffen]

Hi Will

Thanks a lot
How should I change "potentiostat_live_plot.py" to plot a classical CV with the potential on the x-axis and the current on the y-axis?

Best Regards
Steffen

[Will Dickson]

@Steffen

Here is a modified version which plots the classical CV type plot

from __future__ import print_function
import json
import threading
import Queue
import matplotlib.pyplot as plt
import potentiostat as ps


class PotentiostatLivePlot(ps.Potentiostat):
    """
    Custom version of potentiostat with displays a live plot when running a test.
    """

    def __init__(self, port, timeout=10.0, debug=False):
        super(PotentiostatLivePlot,self).__init__(port,timeout=timeout,debug=debug)
        self.data_queue = Queue.Queue()


    def get_default_volt_lims(self):
        volt_range = self.get_volt_range()
        volt_range_value = float(volt_range[:-1])
        volt_lims = -volt_range_value, volt_range_value
        return volt_lims


    def get_default_curr_lims(self):
        curr_range = self.get_curr_range()
        curr_range_value = float(curr_range[:-2])
        curr_lims = -curr_range_value, curr_range_value
        return curr_lims


    def get_curr_unit(self):
        curr_range = self.get_curr_range()
        curr_unit = curr_range[-2:]
        return curr_unit


    def receive_data(self,timeunit):
        while True:
            # Get json data from the device
            sample_json = self.readline()
            sample_json = sample_json.strip()
            try:
                sample_dict = json.loads(sample_json.decode())
            except ValueError:
                continue
            # Put new values in data queue
            if len(sample_dict) > 0:
                tval = sample_dict[ps.TimeKey]*ps.TimeUnitToScale[timeunit]
                volt = sample_dict[ps.VoltKey]
                curr = sample_dict[ps.CurrKey]
                self.data_queue.put({'tval': tval, 'volt': volt, 'curr': curr})
            else:
                self.data_queue.put({})



    def run_test(self, testname, param=None, filename=None, volt_lims=None, curr_lims=None, timeunit='s'):

        if timeunit not in ps.TimeUnitToScale:
            raise RuntimeError('uknown timeunit option {0}'.format(timeunit))

        if volt_lims is None:
            volt_lims = self.get_default_volt_lims()

        if curr_lims is None:
            curr_lims = self.get_default_curr_lims()

        curr_unit = self.get_curr_unit() 

        if param is not None:
            self.set_param(testname,param)

        if filename is not None:
            fid = open(filename,'w')

        t_done = self.get_test_done_time(testname, timeunit=timeunit)

        # Create figure and live plot
        fig = plt.figure()
        plt.ion()

        ax = plt.subplot(111)
        line, = plt.plot([0],[0],'b')
        plt.xlim(*volt_lims)
        plt.ylim(*curr_lims)
        plt.grid('on')
        plt.xlabel('(V)')
        plt.ylabel('({0})'.format(curr_unit))
        plt.title('Potentiostat Live Plot')

        fig.canvas.flush_events()
        plt.pause(0.001)

        # Create lists for storing incoming data
        tval_list = [] 
        volt_list = [] 
        curr_list = []

        # Setup up working thread to collect data from device
        data_worker = threading.Thread(target=self.receive_data,args=(timeunit,))
        data_worker.daemon = True

        # Send command to run the test
        cmd_dict = {
                ps.CommandKey: ps.RunTestCmd, 
                ps.TestKey: testname
                }
        msg_dict = self.send_cmd(cmd_dict)
        self.test_running = True

        data_worker.start()
        done = False

        while not done:

            # Get available data from data queue and add to list
            while self.data_queue.qsize() > 0:
                data = self.data_queue.get(True)
                if not data:
                    done = True # Empty data means run is complete
                    break

                # Get data items and append to lists
                tval = data['tval']
                volt = data['volt']
                curr = data['curr']
                tval_list.append(tval)
                volt_list.append(volt)
                curr_list.append(curr)

                # Write data to file
                if filename is not None:
                    fid.write('{0:1.3f}, {1:1.4f}, {2:1.4f}\n'.format(tval,volt,curr))

                print('{0:1.3f}, {1:1.4f}, {2:1.4f}'.format(tval,volt,curr))

            # Update live plot
            line.set_xdata(volt_list)
            line.set_ydata(curr_list)
            fig.canvas.flush_events()
            plt.pause(0.001)

        self.test_running = False

        if filename is not None:
            fid.close()

        return tval_list, volt_list, curr_list 

 
Note, this version plots the potential and current during the quiet period as well as during the test. However, it should be pretty easy to modify if don't want to show the data during the quiet period. Maybe just discard the incoming data (e.g. don't append it to the lists) until the quiet period is complete.

[pguillem]

Another way would be to raise an async event each time a new value is obtained from the serial?. (it would be triggered "sample_rate" times per second, just as each val is received from the teensy).

That approach could enable any component to register for the event and update itself with each new value, without making the class responisble for plotting. Just thinking out loud.

Right around here:

        if len(sample_dict) > 0:
            tval = sample_dict[ps.TimeKey]*ps.TimeUnitToScale[timeunit]
            volt = sample_dict[ps.VoltKey]
            curr = sample_dict[ps.CurrKey]
            self.data_queue.put({'tval': tval, 'volt': volt, 'curr': curr})

I'll work some code over march and commit to the blog. I´ll also work on some bindings to store long series on DB tables instead of datafile.txt.

Question... can the iorodeo repos be forked with GIT? cheers.

[Will Dickson]

@pguillem

I think separating the plotting from class is a nice idea. And I like the idea of using an event. You might even get away with just the data queue (and no event) ... I hacked together a simple example below

from __future__ import print_function
import json
import threading
import Queue
import potentiostat as ps


class PotentiostatWithQueue(ps.Potentiostat):

    def __init__(self, port, timeout=10.0, debug=False):
        super(PotentiostatWithQueue,self).__init__(port,timeout=timeout,debug=debug)
        self.data_queue = Queue.Queue()

    def receive_data(self,timeunit):

        while self.test_running:
            # Get json data from the device
            sample_json = self.readline()
            sample_json = sample_json.strip()
            try:
                sample_dict = json.loads(sample_json.decode())
            except ValueError:
                continue
            # Put new values in data queue
            if len(sample_dict) > 0:
                tval = sample_dict[ps.TimeKey]*ps.TimeUnitToScale[timeunit]
                volt = sample_dict[ps.VoltKey]
                curr = sample_dict[ps.CurrKey]
                self.data_queue.put({'tval': tval, 'volt': volt, 'curr': curr})
            else:
                self.data_queue.put({})
                self.test_running = False

    def run_test(self, testname, param=None, timeunit='s'):

        if timeunit not in ps.TimeUnitToScale:
            raise RuntimeError('uknown timeunit option {0}'.format(timeunit))
        if param is not None:
            self.set_param(testname,param)

        # Setup up working thread to collect data from device
        data_worker = threading.Thread(target=self.receive_data,args=(timeunit,))
        data_worker.daemon = True
        self.data_queue = Queue.Queue()

        # Send command to run the test
        cmd_dict = {
                ps.CommandKey: ps.RunTestCmd, 
                ps.TestKey: testname
                }
        msg_dict = self.send_cmd(cmd_dict)
        self.test_running = True
        data_worker.start()

 
Then some other part of your program could pull data from the queue as it comes it and you could plot it as you want. Here is a quick and dirty example. I'm assuming the the above snippet is saved as potentiostat_with_queue.py.

from __future__ import print_function
import Queue
from potentiostat_with_queue import PotentiostatWithQueue
import matplotlib.pyplot as plt

volt_lims = (-2.1,2.0)
curr_lims = (-50,50)

dev = PotentiostatWithQueue(port='/dev/ttyACM0')

dev.set_curr_range('100uA')
dev.set_sample_period(10)

name = 'cyclic'
param = {
        'quietValue' : 0.0,
        'quietTime'  : 1000,
        'amplitude'  : 2.0,
        'offset'     : 0.0,
        'period'     : 2000,
        'numCycles'  : 2,
        'shift'      : 0.0,
        }

dev.set_param(name,param)
t_done = dev.get_test_done_time(name)

# Create figure and live plot
fig = plt.figure()
plt.ion()

ax = plt.subplot(111)
line, = plt.plot([0],[0],'b')
plt.xlim(*volt_lims)
plt.ylim(*curr_lims)
plt.grid('on')
plt.xlabel('(V)')
plt.ylabel('(uA)')
plt.title('Potentiostat Live Plot')

fig.canvas.flush_events()
plt.pause(0.001)

# Lists for incoming data
tval_list = [] 
volt_list = [] 
curr_list = []

# Start test
done = False
dev.run_test(name)

while not done:

    # Get data from queue and add store for plotting
    have_new_data = False
    while True:

        try:
            data = dev.data_queue.get(False)
        except Queue.Empty:
            break

        if data:
            tval = data['tval']
            volt = data['volt']
            curr = data['curr']
            print('{0:1.3f}, {1:1.4f}, {2:1.4f}'.format(tval,volt,curr))
            tval_list.append(tval)
            volt_list.append(volt)
            curr_list.append(curr)
            have_new_data = True
        else:
            done = True
            break

    # Update live plot
    if have_new_data:
        line.set_xdata(volt_list)
        line.set_ydata(curr_list)
        fig.canvas.flush_events()
        plt.pause(0.001)

raw_input('done')

 
In a GUI you might use something a bit different than the while loop shown in the simple example above - maybe a timer callback e.g., something which sets the rate at which you want to update your plot. I'm guessing you probably won't always want to update the plot as frequently as data comes in from the device - it would depend on the sample frequency (or period). The sample rate can be anything from very slow up to 1000Hz which might be a bit fast for updating the plot in the GUI (or at least there is not reason to update the plot that fast). So the idea would be to empty the data queue (which could have multiple samples in it) and then update plot in each timer callback.

The repository on bitbucket uses mercurial - so you can't directly fork with git. You might be able to use something like git-remote-hg https://github.com/felipec/git-remote-hg

In the past I've had OK luck with hg-git for mercurial/git for keeping hg and git repos in sync. http://hg-git.github.io/ It can be a little tricky.

You could also use this https://help.github.com/articles/about-github-importer/ However, this is probably only one way trip e.g no going back to hg. So maybe not the best for sharing.

Of course you could also just use mercurial (hg). If you know git then it would probably only take a few minutes to orient yourself with mercurial https://github.com/sympy/sympy/wiki/Git-hg-rosetta-stone

[Steffen]

@pguillem
Good point. I believe it will also be a more logical structure when it comes to GUI development.

@Will-Dickson
Thanks a lot Will. Both scripts works well.

[Rodeo_First]

@Will-Dickson

Nice code. Do you think this would work for chronoamperometry if I change the parameters in the second script?

[Will Dickson]

@Rodeo_First

Yes. However for chronoamperometry you will probably want to change the plot so that it displays current vs time rather than current vs voltage as the voltages are just stepped in chronoamperometry. Even better you could have two real-time plots - one showing the voltage vs time and the other showing the current vs time. This way you could see when the voltages are stepped.

[Rodeo_First]

@Will-Dickson

Thanks again, it was really helpful.