Working with data

This chapter covers how retrieve, manipulate, and analyze datafrom a measurement.

See also:

Obtaining the measured values

When a measurement is started by calling the psCommSimple.Measure method a reference to the instance of that SimpleMeasurement class is returned. The SimpleMeasurement class references instances of the SimpleCurve class in the SimpleMeasurement.SimpleCurveCollection property.

The SimpleCurve class is used to analyze and manipulate the data. Generate new instances of the SimpleCurve class by calling the SimpleMeasurement.NewSimpleCurve method. This method will generate a new SimpleCurve class based on the available DataArrayType in the instance SimpleMeasurement.

For Cyclic Voltammetry or Mux measurements, multiple SimpleCurve objects can be generated.

The SimpleMeasurement.AvailableDataTypes property contains a list of the available DataArrayType in the measurement. To plot a different SimpleCurve, call the SimpleMeasurement.NewSimpleCurve method directly after it has started to generate another SimpleCurve with different units/values.

List<SimpleCurve> chargeCurves = simpleMeasurement.NewSimpleCurve(
    PalmSens.Data.DataArrayType.Time,
    PalmSens.Data.DataArrayType.Charge,
    "Charge/Time"
); (1)
1 Get Charge over Time curves

To access the raw values in the form of an array of doubles use either the SimpleCurve.XAxisValues or the SimpleCurve.YAxisValues properties.

double[] xValues = simpleCurve.XAxisValues;

Get value of the Y Axis a specified index:

double firstYValue = simpleCurve.YAxisValue(0);

Smoothing/Filtering

To smooth a SimpleCurve call one of the SimpleCurve.Smooth methods.

These methods use the Savitsky-Golay algorithm.

The arguments can be either the SmoothLevel enumerator:

SimpleCurve smoothedCurve = simpleCurve.Smooth(SmoothLevel.Medium);

Or an int specifying the window size, i.e. a window of 4 will filter based on the 4 adjacent points in both directions:

SimpleCurve smoothedCurve2 = simpleCurve.Smooth(25);

Baseline Subtraction

Perform a baseline correction by subtracting the SimpleCurve of a baseline measurement from your SimpleCurve using a moving average baseline by calling SimpleCurve.MovingAverageBaseline. To subtract one curve from another, call SimpleCurve.Subtract with the SimpleCurve to subtract as the argument.

SimpleCurve movingAverageBaseline = simpleCurve.MovingAverageBaseline(); (1)
SimpleCurve baselineSubtractedCurve = simpleCurve.Subtract(movingAverageBaseline); (2)
1 Get the moving average baseline curve
2 Get the simple curve with the subtracted baseline

Arithmetic operations

The SimpleCurve class supports other basic operations such as:

  • Addition - SimpleCurve.Add()

  • Subtraction - SimpleCurve.Subtract()

  • Multiplication - SimpleCurve.Multiply()

  • Exponentiation - SimpleCurve.Exponentiate()

  • Differentiation - SimpleCurve.Differentiate()

  • Integration - SimpleCurve.Integrate()

  • Base 10 Logarithm - SimpleCurve.Log10()

  • Average - SimpleCurve.Average()

  • Sum - SimpleCurve.Sum()

  • Minimum - SimpleCurve.Minimum()

  • Maximum - SimpleCurve.Maximum()

Peak and level detection

The SimpleCurve class contains functions for detecting peaks and levels.

There are three peak detection algorithms.

  1. The default algorithm detects peaks using the curve’s derivative.

  2. The shoulder algorithm can detect peaks that are on a slope and missed by the default algorithm.

  3. The LSV/CV algorithm is specifically designed for detecting peaks in Linear Sweep and Cyclic Voltammetry.

The detected peaks are added to SimpleCurve.Peak, an IEnumerable collection of the Peak interface. The Peak interface describes the properties of the peak (i.e. the peak potential, current, height, width, etc.).

Examples of the peak detection are also provided in the Data processing and Peak detection example projects.

Detect peaks with a minimum width of 0.01V, a minimum height of 0.05µA and discard any existing peaks:

activeSimpleCurve.DetectPeaks(
    minPeakWidth=0.01,
    minPeakHeight=0.05,
    clearPeaks=true
);

Now using the the specialized Linear Sweep / Cyclic Voltammetry algorithm:

await activeSimpleCurve.DetectPeaksAsync(
    minPeakWidth=0.01,
    minPeakHeight=0.05,
    clearPeaks=true,
    peakType=PeakTypes.LSVCV
);

Get peak properties from the first peak:

double peakHeight = activeSimpleCurve.Peaks[0].PeakValue;
double peakPotential = activeSimpleCurve.Peaks[0].PeakX;

Level detection works similar to peak detection, except that the results are stored in SimpleCurve.Levels, an IEnumerable collection of the Level class.

Detect levels with a minimum width of 0.5s, a minimum height of 0.05µA, discard any existing levels:

await activeSimpleCurve.DetectLevelsAsync(0.5,0.05,true);

double levelBegin = activeSimpleCurve.Levels[0].LeftX; (1)
double levelEnd = activeSimpleCurve.Levels[0].RightX;
double levelCurrent = activeSimpleCurve.Levels[0].LevelY;
1 Get level properties from the first level

Equivalent circuit fitting

The SimpleCurve.FitEquivalentCircuit() function fits an equivalent circuit model on your data.

The next code example shows how to fit a Randles circuit using the specified inital values and default fit options. It sets the initial values for the solution resistance (series resistor), charge transfer resistance (parallel resistor), and double layer capacitance (parallel capacitor).

FitResult fitResult = await activeSimpleCurve.FitEquivalentCircuit(
    "R(RC)", new double[] { 100, 8000, 1e-8 }
);

double solutionResistance = fitResult.FinalParameters[0]; (1)
double chargeTransferResistance = fitResult.FinalParameters[1];
double doubleLayerCapacitance = fitResult.FinalParameters[2];
1 Get fit results

To change the default fit options and parameters use the following function in combination with the CircuitModel and FitOptionsCircuit classes.

CircuitModel circuitModel = new CircuitModel();
circuitModel.SetEISdata(_activeMeasurement.Measurement.EISdata[0]); (1)

circuitModel.SetCircuit("R(RC)"); (2)

Parameter p = circuitModel.InitialParameters[0]; (3)
p.MaxValue = 1e6;
p.MinValue = 1e4;
p.Value = 1e5; (4)
1 Sets reference to measured data
2 Sets the circuit defined in the CDC code string (Randles circuit)
3 Change bounds and starting value of the solution resistance (in Ω)
4 Set starting value (Ω)

To change the fit options:

FitOptionsCircuit fitOptions = new FitOptionsCircuit();
fitOptions.Model = circuitModel;
fitOptions.RawData = _activeMeasurement.Measurement.EISdata[0]; (1)
fitOptions.MaxIterations = 1000; (2)
fitOptions.MinimumDeltaErrorTerm = 1e-12; (3)
1 Sets reference to measured data
2 The maximum number of iterations, 500 by default
3 The minimum delta in the error term (sum of squares difference between model and data), default is 1e-9

Fit the circuit defined in the CircuitModel and the options specified in the FitOptions:

FitResult fitResult = await activeSimpleCurve.FitEquivalentCircuit(circuitModel, fitOptions);

double solutionResistance = fitResult.FinalParameters[0];
double chargeTransferResistance = fitResult.FinalParameters[1];
double doubleLayerCapacitance = fitResult.FinalParameters[2];

The EIS fit example project also demonstrates how to use the equivalent circuit fitting.