MATLAB: Does “pchip” not return all outputs when used in a MATLAB Function Block

MATLABsimulink

I'm working in MATLAB and Simulink to develop an algorithm that uses "pchip" to do some interpolation. To test the interpolation algorithm, I wrote a MATLAB function that does the processing and worked on that until I was satisfied with the results. I then copied this function into a MATLAB Function Block, so that I could make use of the algorithm in Simulink. When trying to run my model, I get syntax errors when trying to access different fields of the "pchip" output struct, saying those fields no longer exist. Why is this happening?

Best Answer

MATLAB Function Blocks generate code for the contents of the block to integrate it into the Simulink Environment. Accordingly, you need to be careful of Code Generation limitations for any function that gets used in a MATLAB Function Block.

For "pchip", when code is generated from the function, a different implementation is used that changes the structure of the outputs. In order to access the original output fields from the code generation implementation, you must make use of the "unmkpp" function, which will extract all of the information from the output of the code generation implementation.

Related Solutions

MATLAB: Fit a monotonically increasing curve to inexact data

One option is to use the Curve Fitting Tool within MATLAB. You can open this tool using the function "cftool". From this point you have several choices:

1. You can select "Polynomial" and test out some different degrees to see if the fit is good, and if they appear monotonic.

2. You can select "Smoothing Spline" and click "Smoother" until the function appears monotonic.

Note that neither of these methods would strictly enforce monotonicity.

Once you have created a good fit to the data, you can saving the generated fit to your workspace (found under the Fit tab of the CFTool) and then extract the polynomial coefficients of the curve in order to create a symbolic equation.

Another option is to alter the raw data before creating a fit. The following example code is from the comp.soft-sys-matlab newsgroup with thanks to John D'Errico.

This example will alter your raw data ('x' and 'y') to become monotonically increasing via minimal changes:

n = length(x);
C = eye(n);
D = y;
A = diag(ones(n,1),0) - diag(ones(n-1,1),1);
A(end,:) = [];
b = zeros(n-1,1);
opts = optimset('lsqlin');
opts.LargeScale = 'off';
opts.Display = 'none';
yhat = lsqlin(C,D,A,b,[],[],[],[],[],opts);
plot(x,y,'ro',x,yhat,'b-*')

After performing the above calculations, we can consider interpolation methods, as these methods would not produce a monotonic curve from non-monotonic data. From the "cftool", the variable "yhat" is now monotonically increasing. From this point, you can:

1. Run any of the above options

2. Select Interpolation, and then select "Cubic" or "PCHIP".

Note that PCHIP Interpolation on monotonic data will be strictly monotonic. Once again, you will be able to extract the corresponding polynomial coefficients after saving the generated fit to your workspace.

Finally, you could download the following 3rd party toolbox from the MATLAB File Exchange:

http://www.mathworks.com/matlabcentral/fileexchange/24443-slm-shape-language-modeling

Using this toolbox, you can calculate a spline fit to your data, and you can specify that the result is strictly increasing using the following command:

>> pp = slmengine(x,y,'increasing','on','result','pp');

If you would prefer to fit a cubic function to your data, that is equivalent to specifying exactly two knots, one at either end of your data:

>> pp = slmengine(x,y,'knots',[1 10],'increasing','on','result','pp');

From this point, you can then extract the polynomial coefficients with the command "slmpar(pp, 'symabs' )".

MATLAB: Can the ‘fillmissing’ function be used earlier than 2016b? If not, how can I use spline interpolation to fill missing data entries

No, the "fillmissing" function cannot be used earlier than R2016b, as it was introduced in R2016b. However, there are workarounds for using spline interpolation to fill missing data entries.

First, it is important to note that while it may be tempting to use "ismissing" to identify the locations of the missing data entries, "ismissing" was actually changed in R2016b. Prior to R2016b, "ismissing" only takes tables as input. This means that workarounds in releases earlier than R2016b cannot use "ismissing", unless you are working specifically with tables.

However, we can circumvent this by using other built-in functions available earlier than R2016b such as "isnan", which identifies the locations of NaNs in numeric arrays. Other similar functions include "isnat", "isundefined", etc.

With this in mind, here is a workaround that demonstrates how one might achieve spline interpolation to fill missing numeric values in a vector: Take the vector of missing values, and use "isnan" to determine the indices of the NaNs. Use the resulting logical array to determine the query points needed for the "spline" function to appropriately interpolate the NaNs. If, for instance, you have a matrix and would like to interpolate across rows, this procedure can be used for each row.

Here is some code that illustrates this workaround:

% initialize vector with NaN data
v = rand(1,20);
v(2) = NaN;
v(13:16) = NaN;
% grab integer indices of v to input to spline, and find where NaNs are
x = 1:length(v);
m = isnan(v);
% x(~m) contains the indices of the non-NaN values
% v(~m) contains the non-NaN values
% x(m) contains the indices of the NaN values, and thus the points at
% which we would like to query the spline interpolator
s = spline(x(~m),v(~m),x(m));
% replace NaN values with interpolated values; plot to see results
v(m) = s;
plot(v)

If you want to achieve other filling methods, the workflow may be similar. For example, for PCHIP interpolation, you could use the "pchip" function instead of the "spline" function. For the other filling options, the manual workaround will differ, but keep in mind that "isnan", or the appropriate such function, will most likely be useful in any of these endeavors, as they provide a logical array containing the indices of the missing values that can be used to index into the original array for quick replacement of values.

Best Answer

Related Solutions

MATLAB: Fit a monotonically increasing curve to inexact data

MATLAB: Can the ‘fillmissing’ function be used earlier than 2016b? If not, how can I use spline interpolation to fill missing data entries

Related Question