Changing file path in a loop
이전 댓글 표시
I want to run a code where inside every folder i have a file and save in each folder. Can anyone help me?
댓글 수: 17
Stephen23
2022년 10월 17일
Note that you should replace all of those TABLE2ARRAYs and the piles of parentheses with simple curly braces.
For example, this:
table2array((data(:,1)))
should simply be:
data{:,1}
Writing simpler code would make it much more likely that code works.
Rik
2022년 10월 18일
P = 'G:\A_CLIMATIC_BASE\2019';
S = dir(fullfile(P,'*','041A0259.xls'))
This is also code that you posted earlier:
S = dir(fullfile(P, '*', '041A0259.xls'))
for k = 1:numel(S)
F = fullfile(S(k).folder,S(k).name);
T = readtable(F);
... do whatever with this table
end
This code will work as intended and will not result in k being a vector.
Based on the code you posted, you want this:
S = dir(fullfile(P, '*', '041A0259.xls'))
for k = 1:numel(S)
F = fullfile(S(k).folder,S(k).name);
T = readtable(F);
data = T;
d = datetime(strcat(table2array((data(:,1))), {' '}, table2array((data(:,2 )))));
DN = table2array((data(:,3 )));
figure(k );
plot(d, DN );
end
Stephen23
2022년 10월 19일
"this part of the code ... worked!"
Which is why I gave you that code: because it works and is much simpler than your approach.
Remember that you can show your apprecaition by accepting the answer that helped you most.
"but I receive "Undefined function or variable 'thisDir' "."
Clearly you did not define THISDIR. Perhaps you intended to use P instead (because P actually exists):
savefig(fullfile(P,'DN.fig'))
Rik
2022년 10월 19일
Or perhaps
savefig(fullfile(S(k).folder,'DN.fig'))
Rik
2022년 10월 19일
You should really consider completing a basic Matlab tutorial. You may consider doing the Onramp tutorial (which is provided for free by Mathworks).
You seem to be mostly guessing when writing code, instead of reading the documentation of each function you're using. That is very inefficient.
What happens when you call figure(k)? If a figure with that number already exists, it will be made the current figure (meanin any subsequent calls to functions like plot and savefig will use it if you don't explicitly specify another target). If that figure does not exist yet, it will be created. So that is why you end up with 365 different figures open at the end of your code.
Is that actually what you want to happen?
Anyway, the subsequent calls to figure don't have any effect. If you want to show only a single plot in each figure, you should omit hold all. You would do well to read the documentation for that function to see what it does.
Don't guess at what functions do. Matlab isn't cellular biology, it is a software program. There actually is documentation. You don't have to throw things at the wall to see what sticks. You can read the documentation and look at the examples.
Stephen23
2022년 10월 19일
@Rik: very good advice, thank you.
"But now to incorporate two more variables (pressure and temperature)."
Your request is not sufficiently clear: do you want to plot them all at once in one axes, sequentially in one axes, as multiple axes in one figure, or in multiple figures? From your code it is hard to tell: HOLD ALL hints that you want them all plotted simultaneously in one axes, but calling SAVEFIG multiple times hints that you want to plot them sequentially using the same axes.
We can't guess what you want, you have to know what you want to do.
Do not add things to your code that serve no purpose: https://www.mathworks.com/matlabcentral/answers/1827298-changing-file-path-in-a-loop-and-save-figures#comment_2418078
Stephen23
2022년 10월 20일
The same advice applies: keep your code simple! The simplest way to plot those three is to do it sequentially using the same axes&figure. Do not create 365 separate figures using the FIGURE command, all you need to do is call PLOT&SAVEFIG three times. Probably something like this:
for k = 1:numel(S)
F = fullfile(S(k).folder,S(k).name);
T = readtable(F);
D = datetime(strcat(T{:,1},'T',T{:,2}));
%
plot(D,T{:,04})
savefig(fullfile(S(k).folder,'Rain.fig'))
%
plot(D,T{:,13})
savefig(fullfile(S(k).folder,'Pressure.fig'))
%
plot(D,T{:,14})
savefig(fullfile(S(k).folder,'Temperature.fig'))
end
Stephen23
2022년 10월 20일
"but I tried to specify D vector as below with no result:"
As I have not idea what the date text looks like, I cannot comment on the suitability of the InputFormat.
Note that a name-value input consists of one name and one input value (not three input values), so your second syntax will not work (most likely throwing an error about unknown argument):
Stephen23
2022년 10월 20일
" The first collumn is date (2019-01-01) and the second is time (00:00)."
If the times do not have seconds, then do not specify the seconds in the format string.
Jan
2022년 10월 21일
D = datetime(strcat(T{:,1},' ',T{:,2}), 'InputFormat','yyyy/MM/dd',' ', 'HH:mm');
% ^^^ ???
% Maybe you mean:
D = datetime(strcat(T{:,1},' ',T{:,2}), 'InputFormat','yyyy/MM/dd HH:mm');
Stephen23
2022년 10월 22일
Original question from Sirius8 retrieved from Internet Archive:
Changing file path in a loop and save figures
I want to run the code below and change 365 folders, inside every folder i have an excel file (data.xls) and i want to save in each folder a .fig. Right now it only saves the .fig in the first folder. I mean every time it overwrites the previous .fig with the new one. Can anyone help me?
BasePath = 'D:\A_CLIMATIC_BASE\2019';
% Warn user if there is no such folder.
if ~exist(BasePath, 'dir')
message = sprintf('This folder does not exist:\n%s', BasePath);
uiwait(errordlg(message));
return;
end
% Get a list of all files, including folders.
DirList = dir(BasePath);
% Extract only the folders, not regular files.
DirList = DirList([DirList.isdir]); % Folders only
% Warn user if there are no subfolders.
if isempty(DirList)
message = sprintf('This folder does not contain any subfolders:\n%s', BasePath);
uiwait(errordlg(message));
return;
end
% Count the number of subfolders.
numberOfFolders = numel(DirList);
% Loop over all subfolders, processing each one.
for k = 1 : numberOfFolders
thisDir = fullfile(BasePath, DirList(k).name);
fprintf('Processing folder %d of %d: %s\n', k, numberOfFolders, thisDir);
filename='data.xls';
data = readtable(filename);
d = datetime(strcat(table2array((data(:,1))), {' '}, table2array((data(:,2)))));
DN = table2array((data(:,3)));
figure(k);
plot(d, DN);
savefig('DN.fig')
end
Rik
2022년 10월 23일
Why did you delete so many of your comments and edit away your question?
Ancalagon8
2022년 10월 23일
Rik
2022년 11월 9일
Editing away your question is considered rude. You have just decided that the help you received should not benefit others. What give you the right to make that decision?
You must not have been aware of these implication, but you can easily revert the question back to the original state and never do something like this in the future.
Jan
2022년 11월 10일
@Ancalagon8: Please do not delete essential parts of the question. Afterwards the time spent for writing the answers is wasted. If all users do this, this forum would be useless. Therefore your "cleaning" is counter-productive and will reduce the interest to help you in the future.
답변 (2개)
Simon Chan
2022년 10월 15일
Try this to indicate the entire path:
savefig(fullfile(thisDir,'DN.fig'))
댓글 수: 5
Jan
2022년 10월 16일
thisDir = fullfile(BasePath, DirList(k).name);
filename='data.xls';
data = readtable(filename);
Use the folder also:
data = readtable(fullfile(thisDir, 'data.xls'));
Or does the file exist, but is not a valid Excel file?
"I also noticed that the above code runs smoothly inside the first folder 'G:\A_CLIMATIC_BASE\2019\01-01-2019\041A0259.xls'. When i try to run from the initial folder (2019) i receive this error...."
Compare the filepaths:
'G:\A_CLIMATIC_BASE\2019\01-01-2019\041A0259.xls' % what you write in your comment
'G:\A_CLIMATIC_BASE\2019\041A0259.xls' % what you show in your error message
Stephen23
2022년 10월 17일
"I want to run the script from 'G:\A_CLIMATIC_BASE\2019' ..."
You are using absolute filenames, so it should be irrelevant where you run that script from.
"...and automatically find each excel file in every folder (01-01-2019\041A0259.xls, 02-01-2019\041A0259.xls.......31-12-2019\041A0259.xls)"
If that is your actual goal, a better approach would be to use DIR to do more of the heavy lifting, e.g.:
P = 'G:\A_CLIMATIC_BASE\2019';
S = dir(fullfile(P,'*','041A0259.xls'))
Rik
2022년 10월 17일
Why did you modify the code like that? The dir function and readtable functions do completely different things.
You should use dir to determine the name of your file, and then provide readtable with the full path and file name.
"I tried..."
Why did you stick DIR inside the loop like that? The entire point of DIR is to get a list of any files that match the specified name, it will also automatically detect all folders. The code I gave you replaces all of your (complex, non-working) code by using DIR more effectively.
"There is a typo in Stephens suggestion:"
No, I specifically did not want a recursive search. One star will match any subfolder, matching the OPs description.
As I wrote earlier, you need to get DIR to more of the work for you. It is simpler and much more efficient when you let DIR do as much as possible matching names and folders. I will demonstrate this here on the forum, but you should be able to adapt this to your own files.
First lets create some fake data files in subfolders:
for k = 'A':'D' % DO NOT COPY
mkdir(k) % DO NOT COPY
writematrix(rand(3,3),fullfile(k,'test.txt')) % DO NOT COPY
end % DO NOT COPY
clearvars % DO NOT COPY
Now lets use DIR to get a structure array of those files in all subfolders:
P = '.'; % absolute/relative path to where the subfolders are.
S = dir(fullfile(P,'*','test.txt')); % one line is much simpler than your code.
Note how a single asterisk is sufficient to match subfolders (without recursion). We can check what DIR found:
{S.name} % view check the filenames % DO NOT COPY
{S.folder} % view the corresponding subfolder names % DO NOT COPY
And then all you need to do is loop over those files:
for k = 1:numel(S)
F = fullfile(S(k).folder,S(k).name);
... do whatever with this filename
end
댓글 수: 8
Rik
2022년 10월 18일
That part was to create example files, so you can skip that part of the code anyway.
And why did you put instead of *? There is a difference between the two. Are you aware of the difference? Do you understand what each does?
Stephen23
2022년 10월 18일
"So I should skip that part of code but the error remains."
I commented the parts of my answer that are just for testing the code on this forum. You got that error because you copied everything and ignored what I wrote.
load(websave('S.mat','https://www.mathworks.com/matlabcentral/answers/uploaded_files/1160353/S.mat'));
S(1)
Perhaps readtable has trouble with paths containing rich text. I don't see any indication why readtable(F) would not work as expected.
You could consider using xlsread instead, given that you're on an older release.
""But still cannot use readtable."
Why not? What is stopping you from calling READTABLE inside the loop?:
for k = 1:numel(S)
F = fullfile(S(k).folder,S(k).name);
T = readtable(F);
... do whatever with this table
end
"I attach the S (365x1) value"
Thank you, so far it looks exactly as expected:
tmp = load('S.mat');
S = tmp.S
{S.name}
{S.folder}
So far my code worked as expected: it clearly matched 365 files in the subfolders, as the two cell arrays show.
Stephen23
2022년 10월 18일
What is the value of k after the error is thrown?
tmp = load('k.mat')
tmp.k
There is no obvious way that you would get k = 1x365 vector with the code that I gave you: the fact that k is a vector tells us that you are doing something different from what I showed you. Perhaps you did not write the FOR loop correctly... but without seeing your actual code that is just a guess.
If you want further help, please post the exact and unmodified code that you are using (by uploading if it is large) and also upload a mat file containing all workspace variables after the error occurs.
Rik
2022년 10월 24일
I don't see what confidential information might be encoded in the path. You published it yourself under a creative commons license when you posted the mat file. I have therefore removed your flag.
카테고리
도움말 센터 및 File Exchange에서 File Operations에 대해 자세히 알아보기
2022년 10월 15일
2025년 1월 6일
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
