How to find derivative of CSV data plot?

I have a csv file with uniform points but i would like to write MATLAB code for obtain the derivative of the plotted function(unknown) with time. How can i do this?

if true
  % code
end

Answer 1

[Star Strider]

It is easiest to use the gradient (link) function:

[D,S,R] = xlsread('Uniform_points.csv');
t = D(:,1);
f = D(:,2);

dt = mean(diff(t));
df = gradient(f,dt);

figure
subplot(2,1,1)
plot(t, f)
grid
title(S{2},'Interpreter','none')
xlabel(S{1})
subplot(2,1,2)
plot(t, df)
grid
title(['d(',S{2},')/dt'],'Interpreter','none')
xlabel(S{1})

[jasper kansiime]

What would the code be if the points in the csv file were NON UNIFORM? what would change?

[Star Strider]

The work-around I use for that (using the notation in my Answer) is:

dfdt = gradient(f)./gradient(t);

It seems to work well enough, since I’ve successfully used the same approach in other situations.

[jasper kansiime]

okay looks good as well. I would just like to understand what " ./ " means in the line " * dfdt = gradient(f)./gradient(t);* "

[Star Strider]

That creates element-wise division. The ‘.’ in front of multiplication, division, and exponentiation operators makes them element-wise operators. (The exception is the transpose operator ('), where the ‘.’ converts it from a complex-conjugate transpose to an ‘ordinary’ transpose.)

See: Array vs. Matrix Operations (link) for a full discussion.

[jasper kansiime]

Thanks for the discussion link.It clarified things alot.I have another question

In the above code you have" plot(t, f)" where "t" represents time axis. How can i scale the current time axis(0-1second) into an X-axis from(0- 0.4seconds) in steps of 0.04 seconds?

[Star Strider]

To simply alter the plot x-tick labels, this works:

t_new = 0:0.04:0.4;

figure
plot(t, f)
xt = get(gca, 'XTick');
xtv = linspace(min(xt), max(xt), numel(t_new));
set(gca, 'XTick', xtv, 'XTickLabel',t_new)

I generalised it, so it is a bit more complicated than it needs to be for the current values. It will work for other ‘t_new’ vectors without further modification.