Why 'pca' in Matlab doesn't give orthogonal principal components? - matlab

Why using pca in Matlab, I cannot get the orthogonal principal component matrix
For example:
A=[3,1,-1;2,4,0;4,-2,-5;11,22,20];
A =
3 1 -1
2 4 0
4 -2 -5
11 22 20
>> W=pca(A)
W =
0.2367 0.9481 -0.2125
0.6731 -0.3177 -0.6678
0.7006 -0.0150 0.7134
>> PCA=A*W
PCA =
0.6826 2.5415 -2.0186
3.1659 0.6252 -3.0962
-3.9026 4.5028 -3.0812
31.4249 3.1383 -2.7616
Here, every column is a principle component. So,
>> PCA(:,1)'*PCA(:,2)
ans =
84.7625
But the principle component matrix hasn't mutually orthogonal components.
I checked some materials, it said they are not only uncorrelated, but strictly orthogonal. But I can't get the desired result. Can somebody tell me where I went wrong?
Thanks!

You are getting confused between the representation of A in the PCA feature space and the principal components. W are the principal components, and they will indeed be orthogonal.
Check that W(:,1).'*W(:,2) = 5.2040e-17, W(:,1).'*W(:,3) = -1.1102e-16 -- indeed orthogonal
What you are trying to do is to transform the data (i.e. A) in the PCA feature space. You should mean center the data first and then multiply by the principal components as follows.
% A_PCA = (A-repmat(mean(A),4,1))*W
% A more efficient alternative to the above command
A_PCA = bsxfun(#minus,A,mean(A))*W
% verify that it is correct by comparing it with `score` - i.e. the PCA representation
% of A given by MATLAB.
[W, score] = pca(A); % mean centering will occur inside pca
(score-(A-repmat(mean(A),4,1))*W) % elements are of the order of 1e-14, hence equal.

Related

How to interprete the regression plot obtained at the end of neural network regression for multiple outputs?

I have trained my Neural network model using MATLAB NN Toolbox. My network has multiple inputs and multiple outputs, 6 and 7 respectively, to be precise. I would like to clarify few questions based on it:-
The final regression plot showed at the end of the training shows a very good accuracy, R~0.99. However, since I have multiple outputs, I am confused as to which scatter plot does it represent? Shouldn't we have 7 target vs predicted plots for each of the output variable?
According to my knowledge, R^2 is a better method of commenting upon the accuracy of the model, whereas MATLAB reports R in its plot. Do I treat that R as R^2 or should I square the reported R value to obtain R^2.
I have generated the Matlab Script containing weight, bias and activation functions, as a final Result of the training. So shouldn't I be able to simply give my raw data as input and obtain the corresponding predicted output. I gave the exact same training set using the indices Matlab chose for training (to cross check), and plotted the predicted output vs actual output, but the result is not at all good. Definitely, not along the lines of R~0.99. Am I doing anything wrong?
code:
function [y1] = myNeuralNetworkFunction_2(x1)
%MYNEURALNETWORKFUNCTION neural network simulation function.
% X = [torque T_exh lambda t_Spark N EGR];
% Y = [O2R CO2R HC NOX CO lambda_out T_exh2];
% Generated by Neural Network Toolbox function genFunction, 17-Dec-2018 07:13:04.
%
% [y1] = myNeuralNetworkFunction(x1) takes these arguments:
% x = Qx6 matrix, input #1
% and returns:
% y = Qx7 matrix, output #1
% where Q is the number of samples.
%#ok<*RPMT0>
% ===== NEURAL NETWORK CONSTANTS =====
% Input 1
x1_step1_xoffset = [-24;235.248;0.75;-20.678;550;0.799];
x1_step1_gain = [0.00353982300884956;0.00284355877067267;6.26959247648903;0.0275865874012055;0.000366568914956012;0.0533831576137729];
x1_step1_ymin = -1;
% Layer 1
b1 = [1.3808996210168685;-2.0990163849711894;0.9651733083552595;0.27000953282929346;-1.6781835509820286;-1.5110463684800366;-3.6257438832309905;2.1569498669085361;1.9204156230460485;-0.17704342477904209];
IW1_1 = [-0.032892214008082517 -0.55848270745152429 -0.0063993424771670616 -0.56161004933654057 2.7161844536020197 0.46415317073346513;-0.21395624254052176 -3.1570133640176681 0.71972178875396853 -1.9132557838515238 1.3365248285282931 -3.022721627052706;-1.1026780445896862 0.2324603066452392 0.14552308208231421 0.79194435276493658 -0.66254679969168417 0.070353201192052434;-0.017994515838487352 -0.097682677816992206 0.68844109281256027 -0.001684535122025588 0.013605622123872989 0.05810686279306107;0.5853667840629273 -2.9560683084876329 0.56713425120259764 -2.1854386350040116 1.2930115031659106 -2.7133159265497957;0.64316656469750333 -0.63667017646313084 0.50060179040086761 -0.86827897068177973 2.695456517458648 0.16822164719859456;-0.44666821007466739 4.0993786464616679 -0.89370838440321498 3.0445073606237933 -3.3015566360833453 -4.492874075961689;1.8337574137485424 2.6946232855369989 1.1140472073136622 1.6167763205944321 1.8573696127039145 -0.81922672766933646;-0.12561950922781362 3.0711045035224349 -0.6535751823440773 2.0590707752473199 -1.3267693770634292 2.8782780742777794;-0.013438026967107483 -0.025741311825949621 0.45460734966889638 0.045052447491038108 -0.21794568374100454 0.10667240367191703];
% Layer 2
b2 = [-0.96846557414356171;-0.2454718918618051;-0.7331628718025488;-1.0225195290982099;0.50307202195645395;-0.49497234988401961;-0.21817117469133171];
LW2_1 = [-0.97716474643411022 -0.23883775971686808 0.99238069915206006 0.4147649511973347 0.48504023209224734 -0.071372217431684551 0.054177719330469304 -0.25963474838320832 0.27368380212104881 0.063159321947246799;-0.15570858147605909 -0.18816739764334323 -0.3793600124951475 2.3851961990944681 0.38355142531334563 -0.75308427071748985 -0.1280128732536128 -1.361052031781103 0.6021878865831336 -0.24725687748503239;0.076251356114485525 -0.10178293627600112 0.10151304376762409 -0.46453434441403058 0.12114876632815359 0.062856969143306296 -0.0019628163322658364 -0.067809039768745916 0.071731544062023825 0.65700427778446913;0.17887084584125315 0.29122649575978238 0.37255802759192702 1.3684190468992126 0.60936238465090853 0.21955911453674043 0.28477957899364675 -0.051456306721251184 0.6519451272106177 -0.64479205028051967;0.25743349663436799 2.0668075180209979 0.59610776847961111 -3.2609682919282603 1.8824214917530881 0.33542869933904396 0.03604272669356564 -0.013842766338427388 3.8534510207741826 2.2266745660915586;-0.16136175574939746 0.10407287099228898 -0.13902245286490234 0.87616472446622717 -0.027079111747601223 0.024812287505204988 -0.030101536834009103 0.043168268669541855 0.12172932035587079 -0.27074383434206573;0.18714562505165402 0.35267726325386606 -0.029241400610813449 0.53053853235049087 0.58880054832728757 0.047959541165126809 0.16152268183097709 0.23419456403348898 0.83166785128608967 -0.66765237856750781];
% Output 1
y1_step1_ymin = -1;
y1_step1_gain = [0.114200879346771;0.145581598485951;0.000139011547272197;0.000456244862967996;2.05816254143146e-05;5.27704485488127;0.00284355877067267];
y1_step1_xoffset = [-0.045;1.122;2.706;17.108;493.726;0.75;235.248];
% ===== SIMULATION ========
% Dimensions
Q = size(x1,1); % samples
% Input 1
x1 = x1';
xp1 = mapminmax_apply(x1,x1_step1_gain,x1_step1_xoffset,x1_step1_ymin);
% Layer 1
a1 = tansig_apply(repmat(b1,1,Q) + IW1_1*xp1);
% Layer 2
a2 = repmat(b2,1,Q) + LW2_1*a1;
% Output 1
y1 = mapminmax_reverse(a2,y1_step1_gain,y1_step1_xoffset,y1_step1_ymin);
y1 = y1';
end
% ===== MODULE FUNCTIONS ========
% Map Minimum and Maximum Input Processing Function
function y = mapminmax_apply(x,settings_gain,settings_xoffset,settings_ymin)
y = bsxfun(#minus,x,settings_xoffset);
y = bsxfun(#times,y,settings_gain);
y = bsxfun(#plus,y,settings_ymin);
end
% Sigmoid Symmetric Transfer Function
function a = tansig_apply(n)
a = 2 ./ (1 + exp(-2*n)) - 1;
end
% Map Minimum and Maximum Output Reverse-Processing Function
function x = mapminmax_reverse(y,settings_gain,settings_xoffset,settings_ymin)
x = bsxfun(#minus,y,settings_ymin);
x = bsxfun(#rdivide,x,settings_gain);
x = bsxfun(#plus,x,settings_xoffset);
end
The above one is the automatically generated code. The plot which I generated to cross-check the first variable is below:-
% X and Y are input and output - same as above
X_train = X(results.info1.train.indices,:);
y_train = Y(results.info1.train.indices,:);
out_train = myNeuralNetworkFunction_2(X_train);
scatter(y_train(:,1),out_train(:,1))
To answer your question about R: Yes, you should square R to get the R^2 value. In this case, they will be very close since R is very close to 1.
The graphs give the correlation between the estimated and real (target) values. So R is the strenght of the correlation. You can square it to find the R-square.
The graph you draw and matlab gave are not the graph of the same variables. The ranges or scales of the axes are very different.
First of all, is the problem you are trying to solve a regression problem? Or is it a classification problem with 7 classes converted to numeric? I assume this is a classification problem, as you are trying to get the success rate for each class.
As for your first question: According to the literature it is recommended to use the value "All: R". If you want to get the success rate of each of your classes, Precision, Recall, F-measure, FP rate, TP Rate, etc., which are valid in classification problems. values ​​you need to reach. There are many matlab documents for this (help ROC) and you can look at the details. All the values ​​I mentioned and which I think you actually want are obtained from the confusion matrix.
There is a good example of this.
[x,t] = simpleclass_dataset;
net = patternnet(10);
net = train(net,x,t);
y = net(x);
[c,cm,ind,per] = confusion(t,y)
I hope you will see what you want from the "nntraintool" window that appears when you run the code.
Your other questions have already been answered. Alternatively, you can consider using a machine learning algorithm with open source software such as Weka.

Matlab SVM linear binary classification failure

I'm trying to implement a simple SVM linear binary classification in Matlab but I got strange results.
I have two classes g={-1;1} defined by two predictors varX and varY. In fact, varY is enough to classify the dataset in two distinct classes (about varY=0.38) but I will keep varX as random variable since I will need it to other works.
Using the code bellow (adapted from MAtlab examples) I got a wrong classifier. Linear classifier should be closer to an horizontal line about varY=0.38, as we can perceive by ploting 2D points.
It is not displayed the line that should separate two classes
What am I doing wrong?
g(1:14,1)=1;
g(15:26,1)=-1;
m3(:,1)=rand(26,1); %varX
m3(:,2)=[0.4008; 0.3984; 0.4054; 0.4048; 0.4052; 0.4071; 0.4088; 0.4113; 0.4189;
0.4220; 0.4265; 0.4353; 0.4361; 0.4288; 0.3458; 0.3415; 0.3528;
0.3481; 0.3564; 0.3374; 0.3610; 0.3241; 0.3593; 0.3434; 0.3361; 0.3201]; %varY
SVMmodel_testm = fitcsvm(m3,g,'KernelFunction','Linear');
d = 0.005; % Step size of the grid
[x1Grid,x2Grid] = meshgrid(min(m3(:,1)):d:max(m3(:,1)),...
min(m3(:,2)):d:max(m3(:,2)));
xGrid = [x1Grid(:),x2Grid(:)]; % The grid
[~,scores2] = predict(SVMmodel_testm,xGrid); % The scores
figure();
h(1:2)=gscatter(m3(:,1), m3(:,2), g,'br','ox');
hold on
% Support vectors
h(3) = plot(m3(SVMmodel_testm.IsSupportVector,1),m3(SVMmodel_testm.IsSupportVector,2),'ko','MarkerSize',10);
% Decision boundary
contour(x1Grid,x2Grid,reshape(scores2(:,1),size(x1Grid)),[0 0],'k');
xlabel('varX'); ylabel('varY');
set(gca,'Color',[0.5 0.5 0.5]);
hold off
A common problem with SVM or any classification method for that matter is unnormalized data. You have one dimension that spans for 0 to 1 and the other from about 0.3 to 0.4. This causes inbalance between the features. Common practice is to somehow normalize the features, for examply by std. try this code:
g(1:14,1)=1;
g(15:26,1)=-1;
m3(:,1)=rand(26,1); %varX
m3(:,2)=[0.4008; 0.3984; 0.4054; 0.4048; 0.4052; 0.4071; 0.4088; 0.4113; 0.4189;
0.4220; 0.4265; 0.4353; 0.4361; 0.4288; 0.3458; 0.3415; 0.3528;
0.3481; 0.3564; 0.3374; 0.3610; 0.3241; 0.3593; 0.3434; 0.3361; 0.3201]; %varY
m3(:,2) = m3(:,2)./std(m3(:,2));
SVMmodel_testm = fitcsvm(m3,g,'KernelFunction','Linear');
Notice the line before the last.

Vectorizing the solution of a linear equation system in MATLAB

Summary: This question deals with the improvement of an algorithm for the computation of linear regression.
I have a 3D (dlMAT) array representing monochrome photographs of the same scene taken at different exposure times (the vector IT) . Mathematically, every vector along the 3rd dimension of dlMAT represents a separate linear regression problem that needs to be solved. The equation whose coefficients need to be estimated is of the form:
DL = R*IT^P, where DL and IT are obtained experimentally and R and P must be estimated.
The above equation can be transformed into a simple linear model after applying a logarithm:
log(DL) = log(R) + P*log(IT) => y = a + b*x
Presented below is the most "naive" way to solve this system of equations, which essentially involves iterating over all "3rd dimension vectors" and fitting a polynomial of order 1 to (IT,DL(ind1,ind2,:):
%// Define some nominal values:
R = 0.3;
IT = 600:600:3000;
P = 0.97;
%// Impose some believable spatial variations:
pMAT = 0.01*randn(3)+P;
rMAT = 0.1*randn(3)+R;
%// Generate "fake" observation data:
dlMAT = bsxfun(#times,rMAT,bsxfun(#power,permute(IT,[3,1,2]),pMAT));
%// Regression:
sol = cell(size(rMAT)); %// preallocation
for ind1 = 1:size(dlMAT,1)
for ind2 = 1:size(dlMAT,2)
sol{ind1,ind2} = polyfit(log(IT(:)),log(squeeze(dlMAT(ind1,ind2,:))),1);
end
end
fittedP = cellfun(#(x)x(1),sol); %// Estimate of pMAT
fittedR = cellfun(#(x)exp(x(2)),sol); %// Estimate of rMAT
The above approach seems like a good candidate for vectorization, since it does not utilize MATLAB's main strength that is MATrix operations. For this reason, it does not scale very well and takes much longer to execute than I think it should.
There exist alternative ways to perform this computation based on matrix division, as demonstrated here and here, which involve something like this:
sol = [ones(size(x)),log(x)]\log(y);
That is, appending a vector of 1s to the observations, followed by mldivide to solve the equation system.
The main challenge I'm facing is how to adapt my data to the algorithm (or vice versa).
Question #1: How can the matrix-division-based solution be extended to solve the problem presented above (and potentially replace the loops I am using)?
Question #2 (bonus): What is the principle behind this matrix-division-based solution?
The secret ingredient behind the solution that includes matrix division is the Vandermonde matrix. The question discusses a linear problem (linear regression), and those can always be formulated as a matrix problem, which \ (mldivide) can solve in a mean-square error sense‡. Such an algorithm, solving a similar problem, is demonstrated and explained in this answer.
Below is benchmarking code that compares the original solution with two alternatives suggested in chat1, 2 :
function regressionBenchmark(numEl)
clc
if nargin<1, numEl=10; end
%// Define some nominal values:
R = 5;
IT = 600:600:3000;
P = 0.97;
%// Impose some believable spatial variations:
pMAT = 0.01*randn(numEl)+P;
rMAT = 0.1*randn(numEl)+R;
%// Generate "fake" measurement data using the relation "DL = R*IT.^P"
dlMAT = bsxfun(#times,rMAT,bsxfun(#power,permute(IT,[3,1,2]),pMAT));
%% // Method1: loops + polyval
disp('-------------------------------Method 1: loops + polyval')
tic; [fR,fP] = method1(IT,dlMAT); toc;
fprintf(1,'Regression performance:\nR: %d\nP: %d\n',norm(fR-rMAT,1),norm(fP-pMAT,1));
%% // Method2: loops + Vandermonde
disp('-------------------------------Method 2: loops + Vandermonde')
tic; [fR,fP] = method2(IT,dlMAT); toc;
fprintf(1,'Regression performance:\nR: %d\nP: %d\n',norm(fR-rMAT,1),norm(fP-pMAT,1));
%% // Method3: vectorized Vandermonde
disp('-------------------------------Method 3: vectorized Vandermonde')
tic; [fR,fP] = method3(IT,dlMAT); toc;
fprintf(1,'Regression performance:\nR: %d\nP: %d\n',norm(fR-rMAT,1),norm(fP-pMAT,1));
function [fittedR,fittedP] = method1(IT,dlMAT)
sol = cell(size(dlMAT,1),size(dlMAT,2));
for ind1 = 1:size(dlMAT,1)
for ind2 = 1:size(dlMAT,2)
sol{ind1,ind2} = polyfit(log(IT(:)),log(squeeze(dlMAT(ind1,ind2,:))),1);
end
end
fittedR = cellfun(#(x)exp(x(2)),sol);
fittedP = cellfun(#(x)x(1),sol);
function [fittedR,fittedP] = method2(IT,dlMAT)
sol = cell(size(dlMAT,1),size(dlMAT,2));
for ind1 = 1:size(dlMAT,1)
for ind2 = 1:size(dlMAT,2)
sol{ind1,ind2} = flipud([ones(numel(IT),1) log(IT(:))]\log(squeeze(dlMAT(ind1,ind2,:)))).'; %'
end
end
fittedR = cellfun(#(x)exp(x(2)),sol);
fittedP = cellfun(#(x)x(1),sol);
function [fittedR,fittedP] = method3(IT,dlMAT)
N = 1; %// Degree of polynomial
VM = bsxfun(#power, log(IT(:)), 0:N); %// Vandermonde matrix
result = fliplr((VM\log(reshape(dlMAT,[],size(dlMAT,3)).')).');
%// Compressed version:
%// result = fliplr(([ones(numel(IT),1) log(IT(:))]\log(reshape(dlMAT,[],size(dlMAT,3)).')).');
fittedR = exp(real(reshape(result(:,2),size(dlMAT,1),size(dlMAT,2))));
fittedP = real(reshape(result(:,1),size(dlMAT,1),size(dlMAT,2)));
The reason why method 2 can be vectorized into method 3 is essentially that matrix multiplication can be separated by the columns of the second matrix. If A*B produces matrix X, then by definition A*B(:,n) gives X(:,n) for any n. Moving A to the right-hand side with mldivide, this means that the divisions A\X(:,n) can be done in one go for all n with A\X. The same holds for an overdetermined system (linear regression problem), in which there is no exact solution in general, and mldivide finds the matrix that minimizes the mean-square error. In this case too, the operations A\X(:,n) (method 2) can be done in one go for all n with A\X (method 3).
The implications of improving the algorithm when increasing the size of dlMAT can be seen below:
For the case of 500*500 (or 2.5E5) elements, the speedup from Method 1 to Method 3 is about x3500!
It is also interesting to observe the output of profile (here, for the case of 500*500):
Method 1
Method 2
Method 3
From the above it is seen that rearranging the elements via squeeze and flipud takes up about half (!) of the runtime of Method 2. It is also seen that some time is lost on the conversion of the solution from cells to matrices.
Since the 3rd solution avoids all of these pitfalls, as well as the loops altogether (which mostly means re-evaluation of the script on every iteration) - it unsurprisingly results in a considerable speedup.
Notes:
There was very little difference between the "compressed" and the "explicit" versions of Method 3 in favor of the "explicit" version. For this reason it was not included in the comparison.
A solution was attempted where the inputs to Method 3 were gpuArray-ed. This did not provide improved performance (and even somewhat degradaed them), possibly due to wrong implementation, or the overhead associated with copying matrices back and forth between RAM and VRAM.

how to solve a very large overdetermined system of linear equations?

I am doing a project about image processing, and I need to solve the following set of equations:
Nx+Nz*( z(x+1,y)-z(x,y) )=0
Ny+Nz*( z(x+1,y)-z(x,y) )=0
and equations of the boundary (bottom and right side of the image):
Nx+Nz*( z(x,y)-z(x-1,y) )=0
Ny+Nz*( z(x,y)-z(x,y-1) )=0
where Nx,Ny,Nz are the surface normal vectors at the corresponding coordinates and are already determined. Now the problem is that since (x,y) are the coordinates on an image, which typically has a size of say x=300 and y=200. So there are 300x200=60000 unknowns. I rewrite the equations in the form of Mz=b, where M has a size of 120,000x60000, and both z and b are vectors of length 60000. When I solve it using the function in python scipy.linalg.lstsq, I run into memory errors.
I notice that M is very sparse as it only has two non-zero entries of either 1 or -1 in each row. However, I don't know how I can utilize it to solve the problem. Any ideas how I can solve it more efficiently in matlab or python?
In python I find a library that has the lsmr method (as mentioned by someone in the comment). Other than using this algorithm to solve the equation Mx=b, I want to know if I need to store M and B in sparse format as well. Now I just create a very large array with all entries zero in the beginning, then I use a for loop to loop over each pixel and change the corresponding entries to 1 or -1. Then I apply lsmr to solve Mx=b directly. Does it help if I construct the matrix M and b in any one of the sparse format? Right now most of the time spent is in solving Mx=b. Construct the array M,b and doing all previous tasks take negligible time compared to solving Mx=b.
thanks
edit: this is the python code I use to generate matrix M and b. They should be 'correct', but not sure if there are other better ways of rewriting the system of linear equations.
eq_no = 0
for pix in range(total_pix):
row, col = y[pix], x[pix]
if index_array[row,col] >= 0: # confirm (x,y) is inside boundary
# check x-direction
if index_array[row,col+1] >= 0: # (x+1,y) is inside boundary
M[eq_no,pix] = -1
M[eq_no,pix+1] = 1
b[eq_no,0] = -normal_array[row,col,0]/normal_array[row,col,2]
eq_no += 1
# check y-direction
next_pix = index_array[row+1,col]
if next_pix >= 0:
M[eq_no , pix] = -1
M[eq_no , next_pix] = 1
b[eq_no,0] = -normal_array[row,col,1]/normal_array[row,col,2]
eq_no += 1

Getting rank deficient warning when using regress function in MATLAB

I have a dataset comprising of 30 independent variables and I tried performing linear regression in MATLAB R2010b using the regress function.
I get a warning stating that my matrix X is rank deficient to within machine precision.
Now, the coefficients I get after executing this function don't match with the experimental one.
Can anyone please suggest me how to perform the regression analysis for this equation which is comprising of 30 variables?
Going with our discussion, the reason why you are getting that warning is because you have what is known as an underdetermined system. Basically, you have a set of constraints where you have more variables that you want to solve for than the data that is available. One example of an underdetermined system is something like:
x + y + z = 1
x + y + 2z = 3
There are an infinite number of combinations of (x,y,z) that can solve the above system. For example, (x, y, z) = (1, −2, 2), (2, −3, 2), and (3, −4, 2). What rank deficient means in your case is that there is more than one set of regression coefficients that would satisfy the governing equation that would describe the relationship between your input variables and output observations. This is probably why the output of regress isn't matching up with your ground truth regression coefficients. Though it isn't the same answer, do know that the output is one possible answer. By running through regress with your data, this is what I get if I define your observation matrix to be X and your output vector to be Y:
>> format long g;
>> B = regress(Y, X);
>> B
B =
0
0
28321.7264417536
0
35241.9719076362
899.386999172398
-95491.6154990829
-2879.96318251964
-31375.7038251919
5993.52959752106
0
18312.6649115112
0
0
8031.4391233753
27923.2569044728
7716.51932560781
-13621.1638587172
36721.8387047613
80622.0849069525
-114048.707780113
-70838.6034825939
-22843.7931997405
5345.06937207617
0
106542.307940305
-14178.0346010715
-20506.8096166108
-2498.51437396558
6783.3107243113
You can see that there are seven regression coefficients that are equal to 0, which corresponds to 30 - 23 = 7. We have 30 variables and 23 constraints to work with. Be advised that this is not the only possible solution. regress essentially computes the least squared error solution such that sum of residuals of Y - X*B has the least amount of error. This essentially simplifies to:
B = X^(*)*Y
X^(*) is what is known as the pseudo-inverse of the matrix. MATLAB has this available, and it is called pinv. Therefore, if we did:
B = pinv(X)*Y
We get:
B =
44741.6923363563
32972.479220139
-31055.2846404536
-22897.9685877566
28888.7558524005
1146.70695371731
-4002.86163441217
9161.6908044046
-22704.9986509788
5526.10730457192
9161.69080479427
2607.08283489226
2591.21062004404
-31631.9969765197
-5357.85253691504
6025.47661106009
5519.89341411127
-7356.00479046122
-15411.5144034056
49827.6984426955
-26352.0537850382
-11144.2988973666
-14835.9087945295
-121.889618144655
-32355.2405829636
53712.1245333841
-1941.40823106236
-10929.3953469692
-3817.40117809984
2732.64066796307
You see that there are no zero coefficients because pinv finds the solution using the L2-norm, which promotes the "spreading" out of the errors (for a lack of a better term). You can verify that these are correct regression coefficients by doing:
>> Y2 = X*B
Y2 =
16.1491563400241
16.1264219600856
16.525331600049
17.3170318001845
16.7481541301999
17.3266932502295
16.5465094100486
16.5184456100487
16.8428701100165
17.0749421099829
16.7393450000517
17.2993993099419
17.3925811702017
17.3347117202356
17.3362798302375
17.3184486799219
17.1169638102517
17.2813552099096
16.8792925100727
17.2557945601102
17.501873690151
17.6490477001922
17.7733493802508
Similarly, if we used the regression coefficients from regress, so B = regress(Y,X); then doing Y2 = X*B, we get:
Y2 =
16.1491563399927
16.1264219599996
16.5253315999987
17.3170317999969
16.7481541299967
17.3266932499992
16.5465094099978
16.5184456099983
16.8428701099975
17.0749421099985
16.7393449999981
17.2993993099983
17.3925811699993
17.3347117199991
17.3362798299967
17.3184486799987
17.1169638100025
17.281355209999
16.8792925099983
17.2557945599979
17.5018736899983
17.6490476999977
17.7733493799981
There are some slight computational differences, which is to be expected. Similarly, we can also find the answer by using mldivide:
B = X \ Y
B =
0
0
28321.726441712
0
35241.9719075889
899.386999170666
-95491.6154989513
-2879.96318251572
-31375.7038251485
5993.52959751295
0
18312.6649114859
0
0
8031.43912336425
27923.2569044349
7716.51932559712
-13621.1638586983
36721.8387047123
80622.0849068411
-114048.707779954
-70838.6034824987
-22843.7931997086
5345.06937206919
0
106542.307940158
-14178.0346010521
-20506.8096165825
-2498.51437396236
6783.31072430201
You can see that this curiously matches up with what regress gives you. That's because \ is a more smarter operator. Depending on how your matrix is structured, it finds the solution to the system by a different method. I'd like to defer you to the post by Amro that talks about what algorithms mldivide uses when examining the properties of the input matrix being operated on:
How to implement Matlab's mldivide (a.k.a. the backslash operator "\")
What you should take away from this answer is that you can certainly go ahead and use those regression coefficients and they will more or less give you the expected output for each value of Y with each set of inputs for X. However, be warned that those coefficients are not unique. This is apparent as you said that you have ground truth coefficients that don't match up with the output of regress. It isn't matching up because it generated another answer that satisfies the constraints you have provided.
There is more than one answer that can describe that relationship if you have an underdetermined system, as you have seen by my experiments shown above.