MATLAB Answers

How to Run a Code for all files in a folder

10 views (last 30 days)
Hilary Kada
Hilary Kada on 26 Jun 2020
Commented: Hilary Kada on 26 Jun 2020
Hi All!
Am a bit new to Matlab, I have written a code to process csv data and I have a folder full of separate csv's which I would like to run the code on. I'm currently reading each csv into a matrix by referencing the full file name, but I don't really want to change the matrix source for every csv file I have. The final output from my code is a matrix called Graph_Matrix, is it possible to save this matrix for each csv file I run on the code?
function[Graph_Matrix] = transformdata()
%reading in a matrix from the file which I've directly referenced
A = readmatrix('2020_06_25_14_09_19_MetaWear_acc.csv')
%code proceeds
t = A(:,1);
x = A(:,2);
y = A(:,3);
z = A(:,4);
%convert t to appropriate time steps
time_steps = zeros(size(t));
for i = 2:length(t)
time_steps(i) = t(i)-t(i-1)
i = i + 1;
end
%converting time steps to seconds assuming the frequency of the measurement
%is 400Hz
time_step_seconds = time_steps * 0.0025;
time_cumulative = zeros(size(t));
time_cumulative(1) = time_step_seconds (1);
for i = 2:length(t)
time_cumulative(i) = time_step_seconds(i) + time_cumulative(i-1)
i = i + 1;
end
%%
%Converting accelerometer data to accelerations in m/s2
xAcc = x * 9.8;
yAcc = y * 9.8;
zAcc = z * 9.8;
Acc = [xAcc, yAcc, zAcc];
%Calculating the Acceleration Magnitude
AccMag = sqrt(xAcc.^2 + yAcc.^2 + zAcc.^2);
%Converting accelerometer data to velocities in m/s
xVel = zeros(size(xAcc));
yVel= zeros(size(yAcc));
zVel = zeros(size(zAcc));
for i = 2:length(xAcc)
xVel(i) = xAcc(i)*time_step_seconds(i) + xVel(i-1);
yVel(i) = yAcc(i)*time_step_seconds(i) + yVel(i-1);
xVel(i) = zAcc(i)*time_step_seconds(i) + zVel(i-1);
i = i + 1;
end
Vel = [xVel, yVel, zVel];
VelMag = [xVel.^2 + yVel.^2 + zVel.^2];
Graph_Matrix = [time_cumulative, AccMag, VelMag];
end

  0 Comments

Sign in to comment.

Accepted Answer

KSSV
KSSV on 26 Jun 2020
You input the filename to the function transformdata and save the output into cells.
files = dir('*.csv') ;
N = length(files) ;
iwant = cell(N,1) ;
for i = 1:N
thisfile = files(i).name ;
Graph_Matrix = transformdata(thisfile) ;
iwant{i} = Graph_Matrix ;
end
If all the matrices are of same size, you can convert the cell into a 3D matrix using cat.

  3 Comments

Hilary Kada
Hilary Kada on 26 Jun 2020
Thanks! I copied this directly into my code cos i dont fully get it, and when i ran it I got an error for using transformdata(thisfile) because it said I had too many input arguments. When you say I should input the function into transformdata, in the command window do i go transformdata(filename) and then hit enter?
KSSV
KSSV on 26 Jun 2020
Change your atatched function to this:
function[Graph_Matrix] = transformdata(filename)
%reading in a matrix from the file which I've directly referenced
A = readmatrix(filename) ;
%code proceeds
t = A(:,1);
x = A(:,2);
y = A(:,3);
z = A(:,4);
%convert t to appropriate time steps
time_steps = zeros(size(t));
for i = 2:length(t)
time_steps(i) = t(i)-t(i-1)
i = i + 1;
end
%converting time steps to seconds assuming the frequency of the measurement
%is 400Hz
time_step_seconds = time_steps * 0.0025;
time_cumulative = zeros(size(t));
time_cumulative(1) = time_step_seconds (1);
for i = 2:length(t)
time_cumulative(i) = time_step_seconds(i) + time_cumulative(i-1)
i = i + 1;
end
%%
%Converting accelerometer data to accelerations in m/s2
xAcc = x * 9.8;
yAcc = y * 9.8;
zAcc = z * 9.8;
Acc = [xAcc, yAcc, zAcc];
%Calculating the Acceleration Magnitude
AccMag = sqrt(xAcc.^2 + yAcc.^2 + zAcc.^2);
%Converting accelerometer data to velocities in m/s
xVel = zeros(size(xAcc));
yVel= zeros(size(yAcc));
zVel = zeros(size(zAcc));
for i = 2:length(xAcc)
xVel(i) = xAcc(i)*time_step_seconds(i) + xVel(i-1);
yVel(i) = yAcc(i)*time_step_seconds(i) + yVel(i-1);
xVel(i) = zAcc(i)*time_step_seconds(i) + zVel(i-1);
i = i + 1;
end
Vel = [xVel, yVel, zVel];
VelMag = [xVel.^2 + yVel.^2 + zVel.^2];
Graph_Matrix = [time_cumulative, AccMag, VelMag];
end

Sign in to comment.

More Answers (0)