how to get struct array field as a vector and compare the two fields
Mostrar comentarios más antiguos
I have a 1x960 struct array with 3 fields: "IM" , "trialtype", and "memorability"
how can I use the two field "IM" and "memorability" so I can compare them in a graph with "IM" on the x axis and "memorability" on the y
5 comentarios
Walter Roberson
el 15 de Nov. de 2019
What are the data types of the fields, and what are the sizes of the individual entries?
Nadiyah Browning
el 15 de Nov. de 2019
Nadiyah Browning
el 15 de Nov. de 2019
Walter Roberson
el 15 de Nov. de 2019
Editada: Walter Roberson
el 15 de Nov. de 2019
Are they double()? Are they categorical? Are they durations? Are they transfer functions? Are they cell arrays of character vectors that represent dates?
Nadiyah Browning
el 15 de Nov. de 2019
Respuestas (1)
Walter Roberson
el 15 de Nov. de 2019
Under the assumption that IM and memorability are scalars for each entry, then
IM = [YourStructArray.IM];
memorability = [YourStructArray.memorability];
%as an outside observer I cannot assume that the IM entries are in sorted order
[~, idx] = sort(IM);
IM = IM(idx);
memorability = memorability(idx);
plot(IM, memorability)
2 comentarios
Nadiyah Browning
el 15 de Nov. de 2019
Walter Roberson
el 15 de Nov. de 2019
Editada: Walter Roberson
el 15 de Nov. de 2019
"scalar" means a 1 x 1 array, a single value. As compared to the possibility that an IM entry was, for example, a 48 x 64 x 3 RGB image.
If the numbers are random and don't mean anything, then putting them on the x axes needs to be reconsidered, as they provide no position information.
Perhaps what you want is for the memorability values to be plotted in the same order as the index into the array, but for the IM to be used as the label? If so and the IM are numeric, then
IM = [YourStructArray.IM];
memorability = [YourStructArray.memorability];
plot(memorability);
xticks(1:length(IM));
xticklabels(IM);
set(gca, 'XTickLabelRotation', 90);
setting the rotation to 90 is an attempt to deal with the fact that plotting 960 different arbitrary labels on one plot is not going to be readable on anything less than about a 9000 pixel wide display. If the labels turned out to be orderable without gaps (even if the order does not mean anything) then fewer tick labels could be used.
Categorías
Más información sobre Image Arithmetic en Centro de ayuda y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!