Error when attempting to plot ("Error using surf (line 71) Z must be a matrix, not a scalar or vector.")

Hello all, could someone please help me debug this error when I attempt to plot ("Error using surf (line 71) Z must be a matrix, not a scalar or vector."), I'm sure it's something very minor that I'm missing as a newbie.

Thanks.

Xpiv = x;
Ypiv = y;

totalu = zeros(size(Xpiv));
for a = 1:length(velocity_magnitude)
  totalu(:,a) = totalu(:,a) + velocity_magnitude(:,a);
end


pixelToMetricConversion = 1.26; 
fixedPointsM = [137.881,	63.232;124.463,	67.618]; % Teme survey data (gps points)
movingPointsPX = [9.54,3.76;27.42,4.62]; 
movingPointsM = movingPointsPX./pixelToMetricConversion;
mytform = fitgeotrans(movingPointsM, fixedPointsM, 'nonreflectivesimilarity');

Xin = Xpiv./pixelToMetricConversion;
Yin = Ypiv./pixelToMetricConversion;
[Xin,Yin] = transformPointsForward(mytform, Xin (:), Yin(:) );


requiredResolution = 0.5; % in meters
Xin1 = min(Xin):requiredResolution:max(Xin); % equals 0.5m
Yin1 = min(Yin):requiredResolution:max(Yin);
[Xin2,Yin2] = meshgrid(Xin1,Yin1);



normalVel = griddata(Xin, Yin,avu(:),Xin2,Yin2,'nearest');

figure(3);
hold on;
h3 = surf(Xin2,Yin2,normalVel);

[Walter Roberson]

requiredResolution = 0.5; % in meters
Xin1 = min(Xin):requiredResolution:max(Xin); % equals 0.5m
Yin1 = min(Yin):requiredResolution:max(Yin);

Can you be sure that Xin1 and Yin1 will both be vectors with at least two elements? Can you prove that max(Yin)-min(Yin) > requiredDistance? Because if it is not then you would end up with vector Xin2 and Yin2 instead of grids.

[KSSV]

To use surf the dimensions of Xin2, Yin2, normalVel should be matrices and they all should have same dimensions. Check the dimensions.

[Maskus Kit]

@Walter Roberson Thanks a lot. The intention is to end up with grids rather than vectors, not sure how to prove that, any assiatnce to help me with this gridding? Thanks.

[Walter Roberson]

requiredResolution = 0.5; % in meters
minXin = min(Xin);
maxXin = max(Xin);
Xin1 = minXin + (0:ceil((maxXin - minXin)/requiredResolution))*requiredResolution;

Example: 0.2, 0.8, -> ceil((0.8-0.2)/0.5) -> ceil(0.6/0.5) -> ceil(1.2) -> 2 . Then (0:2)*0.5 -> [0, 0.5, 1.0] + 0.2 -> [0.2, 0.7, 1.2]

Also,

normalVel = griddata(Xin, Yin,avu(:),Xin2,Yin2,'nearest', 'extrap');

[Walter Roberson]

load PIVlabExport_28.01.2021.mat


%% this transforms data from PIVlab into a workable format
Xpiv = x;
Ypiv = y;

avu = velocity_magnitude;

%% this transforms the data into the correct co-ordinate sytem that i am working in

pixelToMetricConversion = 1.26; %difference in distance between metric and pixel...find the distance between the co-ordiantes used in each co-ordinate system, using sqrt as shown below. 
% then divde the metric distance by the pixel distance, so 14.11/17.9007 =
% 1.26
% sqrt((9.54-27.42).^2+(3.76-4.62).^2) = 17.9007
% sqrt((137.881-124.463).^2+(63.232-67.618).^2) = 14.1166
fixedPointsM = [137.881,	63.232;124.463,	67.618]; % Teme survey data (gps points)
movingPointsPX = [9.54,3.76;27.42,4.62]; %pixel co-ordinates, or if in PIVlab you have applied a scaling factor (calibration), then it is the PIVlab locations of these points
movingPointsM = movingPointsPX./pixelToMetricConversion;
mytform = fitgeotrans(movingPointsM, fixedPointsM, 'nonreflectivesimilarity');% Reproject the data array so that it is consistent with metric spatial reference used for KLT-IV

Xin = Xpiv./pixelToMetricConversion;
Yin = Ypiv./pixelToMetricConversion;
[Xin,Yin] = transformPointsForward(mytform, Xin (:), Yin(:) );


requiredResolution = 0.5; % in meters
Xin1 = min(Xin):requiredResolution:max(Xin); % equals 0.5m
Yin1 = min(Yin):requiredResolution:max(Yin);
[Xin2,Yin2] = meshgrid(Xin1,Yin1);



normalVel = griddata(Xin, Yin,avu(:),Xin2,Yin2,'nearest');

figure(3);
hold on;
 h3 = surf(Xin2,Yin2,normalVel);

Answer 1

[Walter Roberson]

load PIVlabExport_28.01.2021.mat


%% this transforms data from PIVlab into a workable format
Xpiv = x;
Ypiv = y;

avu = velocity_magnitude;

%% this transforms the data into the correct co-ordinate sytem that i am working in

pixelToMetricConversion = 1.26; %difference in distance between metric and pixel...find the distance between the co-ordiantes used in each co-ordinate system, using sqrt as shown below. 
% then divde the metric distance by the pixel distance, so 14.11/17.9007 =
% 1.26
% sqrt((9.54-27.42).^2+(3.76-4.62).^2) = 17.9007
% sqrt((137.881-124.463).^2+(63.232-67.618).^2) = 14.1166
fixedPointsM = [137.881,	63.232;124.463,	67.618]; % Teme survey data (gps points)
movingPointsPX = [9.54,3.76;27.42,4.62]; %pixel co-ordinates, or if in PIVlab you have applied a scaling factor (calibration), then it is the PIVlab locations of these points
movingPointsM = movingPointsPX./pixelToMetricConversion;
mytform = fitgeotrans(movingPointsM, fixedPointsM, 'nonreflectivesimilarity');% Reproject the data array so that it is consistent with metric spatial reference used for KLT-IV

Xin = Xpiv./pixelToMetricConversion;
Yin = Ypiv./pixelToMetricConversion;
[Xin,Yin] = transformPointsForward(mytform, Xin (:), Yin(:) );


requiredResolution = 0.5; % in meters
Xin1 = min(Xin):requiredResolution:max(Xin); % equals 0.5m
Yin1 = min(Yin):requiredResolution:max(Yin);
[Xin2,Yin2] = meshgrid(Xin1,Yin1);



normalVel = griddata(Xin, Yin,avu(:),Xin2,Yin2,'nearest');

figure(3);
hold on;
 h3 = surf(Xin2,Yin2,normalVel);