dimensional data string

I have an exercise in mathlab .....

load xx18.dat

load yy18.dat

plot(xx18,yy18)

a=mean(xx18)

b=mean(yy18)

c=mean(xx18.^2)

d=mean(yy18.^2)

e=mean(xx18.*yy18)

r=(e-a*b)/sqrt((c-a^2)*(d-b^2))

p=polyfit(xx18,yy18,2)

yt=polyval(p,xx18)

subplot(2,3,1)

plot(xx18,yy18,'r', xx18, yt)

title('Y(X) si Yteoretic(X)')

eroare= yy18-yt

subplot(2,3,2)

plot( eroare,xx18)

title('graficul erorii in raport cu X')

subplot(2,3,3)

plot(eroare,yt)

title('graficul erorii in raport cu Yteoretic')

subplot

load xx18.dat

load yy18.dat

plot(xx18,yy18)

a=mean(xx18)

b=mean(yy18)

c=mean(xx18.^2)

d=mean(yy18.^2)

e=mean(xx18.*yy18)

r=(e-a*b)/sqrt((c-a^2)*(d-b^2))

p=polyfit(xx18,yy18,2)

yt=polyval(p,xx18)

subplot(2,3,1)

plot(xx18,yy18,'r', xx18, yt)

title('Y(X) si Yteoretic(X)')

eroare= yy18-yt

subplot(2,3,2) plot( eroare,xx18) title('graficul erorii in raport cu X')

subplot(2,3,3)

plot(eroare,yt)

title('graficul erorii in raport cu Yteoretic')

subplot(2,3,4)

for i=2:length(eroare) hold on plot(eroare(i-1), eroare(i), '*') end

title('graficul punctelor succesive')

subplot(2,3,5)

hist(eroare,10)

title('histograma erorilor')

subplot(2,3,6)

normplot(eroare)

title('graficul probabilitatilor normale')

v=1:length(eroare)

m=polyfit(v,eroare,1)

How do I calculate the slope, where slope is 0 and the location is constant?

and i don't understand those graphs that are generated... 6 graphics ...and what involving them?

[Paulo Silva]

those are good questions... to ask your teacher

[marciuc]

my teacher is not a normal person....you don't understand the situation :D

[Paulo Silva]

I had some teachers like those in my time, good luck

[Jan]

I assume the titles over the diagrams contain some helpful hints... Sometimes asking even not normal persons yields to useful answers. It is always worth to try it.

Answer 1

[Matt Tearle]

polyfit is finding the least-squares best fit to the data for a quadratic polynomial. polyval then evaluates this polynomial at the x data values. Hence yt is a vector of the y values for the fitted (theoretical) model. eroare is therefore the vector of residuals -- ie the difference between the y values predicted by the fitted model and the actual data.

The graphs are showing the data and the fitted model, then various investigations of the residuals, specifically independence and normality.