using cd to navigate in and out of directories
in [Matlab]
|
Prev: jit acceleration doesn't show up in profiler
Next: Can I disable 'timeout' for the instrument control?
From: seedsp77 Seven on 16 Jun 2010 13:29 Hello, I have a question regarding looping with string characters and cd into directories. I'll give an example: On my desktop, I have a folder called "simulations" that contains multiple runs of a simulation. Within this "simulations" folder contains multiple folders that consist of each simulation run. So for example, the directory containing the first simulation run would be: ~/Desktop/simulation/run1 If there are 10 runs, then there would be 10 "run" folders within the "simulation" folder. Now, within each run folder is contained a ".mat" file that contains the data for the simulation run. The ".mat" file has the same name within each run folder, called "data.mat" What I would like to do is to create a loop that can go through each "run" folder and manipulate its "data.mat" file, such as plotting terms contained within it. Because each run folder has the same initial title, and is only distinguished with a number at the end, I wanted to know if there was a way that I could designate this folder through a for loop. For example, I know that this does not have the correct syntax, but I would like to do something like this: ****** clc clear cd ~/Desktop/simulations for i = 1:10 dir = strcat("run",i); cd dir; load 'data.mat' %% Manipulate by plotting values in 'data.mat' cd .. end Can someone please help me to do something like this? Thank you very much.
From: TideMan on 16 Jun 2010 16:11 On Jun 17, 5:29 am, "seedsp77 Seven" <scf...(a)yahoo.com> wrote: > Hello, I have a question regarding looping with string characters and cd into directories. I'll give an example: > > On my desktop, I have a folder called "simulations" that contains multiple runs of a simulation. Within this "simulations" folder contains multiple folders that consist of each simulation run. So for example, the directory containing the first simulation run would be: > > ~/Desktop/simulation/run1 > > If there are 10 runs, then there would be 10 "run" folders within the "simulation" folder. > > Now, within each run folder is contained a ".mat" file that contains the data for the simulation run. The ".mat" file has the same name within each run folder, called "data.mat" > > What I would like to do is to create a loop that can go through each "run" folder and manipulate its "data.mat" file, such as plotting terms contained within it. Because each run folder has the same initial title, and is only distinguished with a number at the end, I wanted to know if there was a way that I could designate this folder through a for loop. > > For example, I know that this does not have the correct syntax, but I would like to do something like this: > > ****** > clc > clear > cd ~/Desktop/simulations > for i = 1:10 > dir = strcat("run",i); > cd dir; > load 'data.mat' > %% Manipulate by plotting values in 'data.mat' > cd .. > end > > Can someone please help me to do something like this? Thank you very much. Another (better?) way to do it is to stay in the present directory and point to the correct place in the filename: matfile=['~/Desktop/simulations/run' num2str(i) '.mat']; load(matfile) BTW, it's BAD practice to use i for indexing because by default it is sqrt(-1). Use ifile or irun or something like that.
From: us on 16 Jun 2010 16:26 TideMan > BTW, it's BAD practice to use i for indexing because by default it is > sqrt(-1). > Use ifile or irun or something like that. that's EXACTLY right - and the very reason why we use this loop-index-var-name ALL the time... a=zeros(2); b=ones(2); for jTideManThisForLoopIndexerVariable=1:2 for iTideManThisForLoopIndexerVariable=1:2 if iTideManThisForLoopIndexerVariable ~= jTideManThisForLoopIndexerVariable a(iTideManThisForLoopIndexerVariable,iTideManThisForLoopIndexerVariable)=... b(jTideManThisForLoopIndexerVariable,jTideManThisForLoopIndexerVariable); else disp(sprintf('matching TideMen: %5d %5d',... iTideManThisForLoopIndexerVariable,... iTideManThisForLoopIndexerVariable)); end end end :-) us
From: TideMan on 16 Jun 2010 19:37
On Jun 17, 8:26 am, "us " <u...(a)neurol.unizh.ch> wrote: > TideMan > > > BTW, it's BAD practice to use i for indexing because by default it is > > sqrt(-1). > > Use ifile or irun or something like that. > > that's EXACTLY right - and the very reason why we use this loop-index-var-name ALL the time... > > a=zeros(2); > b=ones(2); > for jTideManThisForLoopIndexerVariable=1:2 > for iTideManThisForLoopIndexerVariable=1:2 > if iTideManThisForLoopIndexerVariable ~= jTideManThisForLoopIndexerVariable > a(iTideManThisForLoopIndexerVariable,iTideManThisForLoopIndexerVariable)=... > b(jTideManThisForLoopIndexerVariable,jTideManThisForLoopIndexerVariable); > else > disp(sprintf('matching TideMen: %5d %5d',... > iTideManThisForLoopIndexerVariable,... > iTideManThisForLoopIndexerVariable)); > end > end > end > > :-) > us Oh, us! You exaggerate. If you must use single letters, what's wrong with k, m or n? I never use lower(L) because it's easily mixed up with ones(1,1), just like upper(o) and zeros(1,1). |