I'm using this code to load the ECG-ID database into MATLAB:
%% Initialization
clear all; close all; clc
%% read files from folder A
% Specify the folder where the files live.
myFolder = 'Databases\ECG_ID';
% Check to make sure that folder actually exists. Warn user if it doesn't.
if ~isfolder(myFolder)
errorMessage = sprintf('Error: The following folder does not exist:\n%s\nPlease specify a new folder.', myFolder;)
uiwait(warndlg(errorMessage);)
myFolder = uigetdir(; % Ask for a new one.)
if myFolder == 0
% User clicked Cancel
return;
end
end
% Get a list of all files in the folder with the desired file name pattern.
filePattern = fullfile(myFolder, '**/rec_*'; % Change to whatever pattern you need.)
theFiles = dir(filePattern;)
for k = 1 : length(theFiles)
baseFileName = theFiles(k.name;)
fullFileName = fullfile(theFiles(k.folder, baseFileName);)
fprintf(1, 'Now reading %s\n', fullFileName;)
% Now do whatever you want with this file name,
% such as reading it in as an image array with imread()
[sig, Fs, tm] = rdsamp(fullFileName, [1],[],[],[],1;)
end
But I keep getting this error message:
Now reading C:\Users\******\Documents\MATLAB\Databases\ECG_ID\Person_01\rec_1.atr
Error using rdsamp (line 203)
Could not find record: C:\Users\******\Documents\MATLAB\Databases\ECG_ID\Person_01\rec_1.atr. Search path is set to: '.
C:\Users\******\Documents\MATLAB\mcode\..\database\ http://physionet.org/physiobank/database/'
I can successfully load one signal at a time (but I can't load the entire database using the above code) using this command:
[sig, Fs, tm] = rdsamp('Databases\ECG_ID\Person_01\rec_1');
How do I solve this problem? How can I load all the files in MATLAB?
Thanks in advance.
I have TWO large, CSV files (around 1GB). They both share relation between each other (ID is lets say like a foreign key). Structure is simple, line by line but CSV cells with a line break in the value string can appear
37373;"SOMETXT-CRCF or other other line break-";3838383;"sasa ssss"
One file is P file and other is T file. T is like 70% size of the P file (P > T). I must cut them to smaller parts since they are to big for the program I have to import them... I can not simply use split -l 100000 since I will loose ID=ID relations which must be preserved! Relation can be 1:1, 2:3, 4:6 or 1:5. So stupid file splitting is no option, we must check the place where we create a new file. This is example with simplified CSV structure and a place where I want the file to be cut (and the lines above go to separate P|T__00x file and we continue till P or T ends). Lines are sorted in both files, so no need to search for IDs across whole file!
File "P" (empty lines for clearness):
CSV_FILE_P;HEADER;GOES;HERE
564788402;1;1;"^";"01"
564788402;2;1;"^";"01"
564788402;3;1;"^";"01"
575438286;1;1;"^";"01"
575438286;1;1;"^";"01"
575438286;2;1;"37145859"
575438286;2;1;"37145859"
575438286;3;1;"37145859"
575438286;3;1;"37145859"
575439636;1;1;"^"
575439636;1;1;"^"
# lets say ~100k line limit of file P is somewhere here and no more 575439636 ID lines , so we cut.
575440718;1;1;"^"
575440718;1;1;"^"
575440718;2;1;"10943890"
575440718;2;1;"10943890"
575440718;3;1;"10943890"
575440718;3;1;"10943890"
575441229;1;1;"^";"01"
575441229;1;1;"^";"01"
575441229;2;1;"4146986"
575441229;2;1;"4146986"
575441229;3;1;"4146986"
575441229;3;1;"4146986"
File T (empty lines for clearness)
CSV_FILE_T;HEADER;GOES;HERE
564788402;4030000;1;"0204"
575438286;6102000;1;"0408"
575438286;6102000;0;"0408"
575439636;7044010;1;"0408"
575439636;7044010;0;"0408"
# we must cut here since bigger file "P" 100k limit has been reached
# and we end here because 575439636 ID lines are over.
575440718;6063000;1;"0408"
575440718;6063000;0;"0408"
575441229;8001001;0;"0408"
575441229;8001001;1;"0408"
Can you please help splitting those two files into many 100 000 (or so) lines separate files T_001 and corresponding P_001 file and so on? So ID matches between file parts. I believe awk will be the best tool but I have not got much experience in this field. And the last thing - CSV header should be preserved in each of the files.
I have powerful AIX machine to cope with that (linux also possible since AIX commands are limited sometimes)
You can parse the beginning IDs with awk and then check to see if the current ID is the same as the last one. Only when it is different are you allowed close the current output file and open a new one. At that point record the ID for tracking the next file. You can track this id in a text file or in memory. I've done it in memory but with big files like this you could run into trouble. It's easier to keep track in memory than opening multiple files and reading from them.
Then you just need to distinguish between the first file (output and recording) and the second file (output and using the prerecorded data).
The code does a very brute force check on the possibility of a CRLF in a field - if the line does not begin with what looks like an ID, then it outputs the line and does no further testing on it. Which is a problem if the CRLF is followed immediately by a number and semicolon! This might be unlikely though...
Run with: gawk -f parser.awk P T
I don't promise this works!
BEGIN {
MAXLINES = 100000
block = 0
trackprevious = 0
}
FNR == 1 {
# First line is CSV header
csvheader = $0
if (FILENAME == "-")
{
_error = 1
print "Error: Need filename on command line"
exit 1
}
if (trackprevious)
{
_error = 1
print "Only one file can track another"
exit 1
}
if (block >= 1)
{
# New file - track previous output...
close(outputname)
Tracking[block] = idval
print "Output for " FILENAME " tracks previous file"
trackprevious = 1
}
else
{
print "Chunking output (" MAXLINES ") for " FILENAME
}
linecount = 0
idval = 0
block = 1
outputprefix = FILENAME "_block"
outputname = sprintf("%s_%03d", outputprefix, block)
print csvheader > outputname
next
}
/^[0-9]+;/ {
linecount++
newidval = $0
sub(/;.*$/, "", newidval)
newidval = newidval + 0 # make a number
startnewfile = 0
if (trackprevious && (idval != newidval) && (idval == Tracking[block]))
{
startnewfile = 1
}
else if (!trackprevious && (idval != newidval) && (linecount > MAXLINES))
{
# Last ID value found before new file:
Tracking[block] = idval
startnewfile = 1
}
if (startnewfile)
{
close(outputname)
block++
outputname = sprintf("%s_%03d", outputprefix, block)
print csvheader > outputname
linecount = 1
}
print $0 > outputname
idval = newidval
next
}
{
linecount++
print $0 > outputname
}
Ok. Many trouble shooting hours ...and many error "dings" later, I'm still having the same problem. Due to my beginner skills I'm having problems achieving the following segment of my project:
I will be as detailed as possible so I can hopefully nail it this time:
On my computer i have a folder C:\data which contains many different subfolders.
The subfolders are named by dates in a MMDDYY fashion. For example "040312"
In each subfolder are excel files named after Baseball teams. each subfolder may contain a different combination of xls files.
I am trying to write code that achieves the following objectives:
1.) Loops through all the subfolders of the C:\data folder looking for xls files that have the filenames: Angles.xls, Diamondbacks.xls, etc.
2.) If the files are found in each subfolder import the spreadsheet data and generate a plot of the data titled "Score" and "Allow".
3.) If the file is not found any given subfolder skip and continue to the next file to be located.
4.)Then save the generated plot in the same folder that the spreadsheet was imported from as a .fig and a .bmp file.
I've gotten hints to use various functions like: genpath, dir, but the code I've been fumbling through isn't able to achieve my goals.
a) the script doesn't import the excel files from all the subfolders
b) the script wont save the .fig or .bmp file in the associated subfolder
Here is the code I have been fumbling through:
%I know all of this is wrong wrong wrong. Please help to adjust my code to %achieve the objectives outlined above!
addpath(genpath('c:\data'))
folder = 'c:\data';
subdirs = dir(folder);
subdirs(~[subdirs.isdir]) = [] ;
numberOfFolders = length(subdirs);
if numberOfFolders <= 0
uiwait(warndlg('Number of folders = 0!'))
end
wantedfiles = {'Angels' 'Diamondbacks' 'Orioles' 'Royals' 'Yankees' 'Mets' 'Giants'};
for K = 1 : numberOfFolders
thissubdir = subdirs(K).name;
if strcmp(thissubdir, '.') || strcmp(thissubdir, '..')
continue;
end
subdirpath = [folder '\' thissubdir];
for L = 1 : length(wantedfiles)
for wantedfiles = {'Angels' 'Diamondbacks' 'Orioles' 'Royals' 'Yankees' 'Mets' 'Giants'};
folder = '';
fileToRead1 = [wantedfiles{1} '.xls'];
sheetName='Sheet1';
if exist(fileToRead1, 'file') == 0
% File does not exist
% Skip to bottom of loop and continue with the loop
continue;
end
%This is to import the data and organize it
% All of this code I had auto-generated from importing files manually
[numbers, strings, raw] = xlsread(fileToRead1, sheetName);
if ~isempty(numbers)
newData1.data = numbers;
end
if ~isempty(strings) && ~isempty(numbers)
[strRows, strCols] = size(strings);
[numRows, numCols] = size(numbers);
likelyRow = size(raw,1) - numRows;
% Break the data up into a new structure with one field per column.
if strCols == numCols && likelyRow > 0 && strRows >= likelyRow
newData1.colheaders = strings(likelyRow, :);
end
end
% Create new variables in the base workspace from those fields.
for i = 1:size(newData1.colheaders, 2)
assignin('base', genvarname(newData1.colheaders{i}), newData1.data(:,i));
end
% Now I execute the plotting of data
subplot (2,1,1), plot(Score,Allow)
title([wantedfiles{1} 'Testing to see if it works']);
subplot (2,1,2), plot(Allow,Score)
title('Well, did it?');
% here I save the generated plots, but they don't save where I want them to
saveas(gcf,[wantedfiles{1} ' did it work.fig']);
saveas(gcf,[wantedfiles{1} ' did it work.bmp']);
end
end
end
%At the end of the script I still was unable to loop over the files that I wanted
rmpath(genpath('c:\data'));
Some of you may notice I'm already back with the same painful code already. I'm not sure if the other question is still open or not once I accept an answer.
Now the problem is a little simpler. I found some code that checked for pangrams. It use to be def pangram?('sentence') but I needed line to go in there so I tried changing it to def pangram?(line). It doesn't seem to mesh well with my coding style and doesn't work. I tried to use .contain('a' . . 'z') to check for a pangram but someone I know tried that and it didn't work. Also google isn't much help either.
Any ideas for how I could check for pangrams in an if stmt?
# To change this template, choose Tools | Templates
# and open the template in the editor
# This program reads a file line by line,
#separating lines by writing into certain text files.
#PPQ - Pangrams, Palindromes, and Quotes
class PPQ
def pangram?(line)
unused_letters = ('a'..'z').to_a - line.downcase.chars.to_a
unused_letters.empty?
end
def categorize
file_pangram = File.new('pangram.txt', 'w')
file_palindrome = File.new('palindrome.txt', 'w')
file_quotes = File.new('quotes.txt','w')
File.open('ruby1.txt','r') do |file|
while line = file.gets
if(line.reverse == line)
file_palindrome.write line
elsif(pangram?(line)== true)
file_pangram.write line
else
file_quotes.write line
end
end
end
file.close
file_pangram.close
file_palindrome.close
file_quotes.close
end
end
my_ruby_assignment = PPQ.new
my_ruby_assignment.categorize
I'm partial to simpler syntax, something like
def pangram?(line)
('a'..'z').all? { |word| line.downcase.include? (word) }
end
if pangram?(line) then file_pangram.write line end
def pangram?(string)
str = string.chars.map(&:downcase)
letters =('a'..'z').to_a
result = true
letters.each do |l|
if !(str.include? l.downcase)
result = false
break
end
end
result
end
My customer is sending TDM/TDX files captured in National Instruments Diadem, which I haven't got. I'm looking for a way to convert the files into .CSV, XLS or .MAT files for analysis in Matlab (without using Diadem or Diadem DLLs!)
The format consists of a well structured XML file (.TDM) and a binary (.TDX), with the .TDM defining how fields are packed as bits in the binary TDX. I'd like to read the files (for use in Matlab and other environments). Does anyone have a general purpose tool or conversion script in for instance Python or Perl (not using the NI DLL's) or directly in Matlab?
I've looked into buying the tool, but didn't like it for anything other than one-time conversion to a compatible file format.
Thanks!
I know this is a little late, but I have a simple library to read TDM/TDX files in Python. It works by parsing the TDM file to figure out the data type, then using NumPy.memmap to open the TDX file. It can then be used like a standard NumPy array. The code is pretty simple, so you could probably implement something similar in Matlab.
Here's the link: https://bitbucket.org/joshayers/tdm_loader
Hope that helps.
Maybe a little too late, but I think there is a simple way to get the data from TDM files: NI provides plug-ins for reading TDM files into Excel and OpenOffice Calc. Having the data in one of these programs you could use the CSV export. Search google for "tdm excel" or "tdm openoffice".
Hope this helps...
Gemue
The following script can convert all variables into 'variable' struct.
CurrDirectory = '...//'; % Path to current directory
fileNametdx = '.../utility/'; % Path to TDX file
%%
% Data type conversion
Dtype.eInt8Usi='int8';
Dtype.eInt16Usi='int16';
Dtype.eInt32Usi='int32';
Dtype.eInt64Usi='int64';
Dtype.eUInt8Usi='uint8';
Dtype.eUInt16Usi='uint16';
Dtype.eUInt32Usi='uint32';
Dtype.eUInt64Usi='uint64';
Dtype.eFloat32Usi='single';
Dtype.eFloat64Usi='double';
%% Read .tdx file Name
wb=waitbar(0,'Reading *.tdx Files');
fileNameTDM = strrep(fileNametdx,'.tdx','.TDM');
%% Read .TDM
tdm=xml2struct(fileNameTDM);
for i=1:numel(tdm.usi_colon_tdm.usi_colon_data.tdm_channel)
waitbar((1/numel(tdm.usi_colon_tdm.usi_colon_data.tdm_channel))*i,wb,['File ' fileNametdx ' conversion started']);
s1=strsplit(string(tdm.usi_colon_tdm.usi_colon_data.tdm_channel{1, i}.local_columns.Text),'"');
usi1=s1(2);
% if condition match untill we get usi2
for j=1:numel(tdm.usi_colon_tdm.usi_colon_data.localcolumn)
usi2=string(tdm.usi_colon_tdm.usi_colon_data.localcolumn{1, j}.Attributes.id);
if usi1==usi2
%take new usi
s2=strsplit(string(tdm.usi_colon_tdm.usi_colon_data.localcolumn{1, j}.values.Text),'"');
new_usi1=s2(2);
w1=strsplit(string(tdm.usi_colon_tdm.usi_colon_data.tdm_channel{1, i}.datatype.Text),'_');
str_1=char(strcat('tdm.usi_colon_tdm.usi_colon_data.',lower(w1(2)),'_sequence'));
str_2=char(strcat('tdm.usi_colon_tdm.usi_colon_data.',lower(w1(2)),'_sequence{1, k}.Attributes.id'));
str_3=char(strcat('tdm.usi_colon_tdm.usi_colon_data.',lower(w1(2)),'_sequence{1, k}.values.Attributes.external'));
str_4=char(strcat('tdm.usi_colon_tdm.usi_colon_data.',lower(w1(2)),'_sequence{1, k}.values'));
for k=1:numel(eval(str_1))
new_usi2=string(eval(str_2));
if new_usi1==new_usi2
if isfield(eval(str_4), 'Attributes')
inc_value1=string(eval(str_3));
for m=1:numel(tdm.usi_colon_tdm.usi_colon_include.file.block)
inc_value2=string(tdm.usi_colon_tdm.usi_colon_include.file.block{1, m}.Attributes.id);
if inc_value1==inc_value2
% offset=round(str2num(tdm.usi_colon_tdm.usi_colon_include.file.block{1, m}.Attributes.byteOffset)/8);
length = round(str2num(tdm.usi_colon_tdm.usi_colon_include.file.block{1, m}.Attributes.length));
offset1=round(str2num(tdm.usi_colon_tdm.usi_colon_include.file.block{1, m}.Attributes.byteOffset));
value_type = tdm.usi_colon_tdm.usi_colon_include.file.block{1, m}.Attributes.valueType;
m = memmapfile(fullfile(CurrDirectory,fileNametdx),'Offset',offset1,'Format',{Dtype.(value_type) [length 1] 'dat'},'Writable',true,'Repeat',1);
dat=m.Data.dat ;
end
end
else
str_5=char(strcat('tdm.usi_colon_tdm.usi_colon_data.',lower(w1(2)),'_sequence{1, k}.values.',char(fieldnames(tdm.usi_colon_tdm.usi_colon_data.string_sequence{1, k}.values))));
dat=eval(str_5)';
end
name_variable = string(tdm.usi_colon_tdm.usi_colon_data.tdm_channel{1, i}.name.Text);
varname = genvarname(char(name_variable));
variable.(varname) = dat;
end
end
end
end
end
waitbar(1,wb,[fileNametdx ' conversion completed']);
pause(1)
close(wb)
delete(fullfile(CurrDirectory,fileNametdx),fullfile(CurrDirectory,fileNameTDM));
%Output Variable is Struct
clearvars -except variable
This script requires following XML parser
function [ s ] = xml2struct( file )
%Convert xml file into a MATLAB structure
% [ s ] = xml2struct( file )
%
% A file containing:
% <XMLname attrib1="Some value">
% <Element>Some text</Element>
% <DifferentElement attrib2="2">Some more text</Element>
% <DifferentElement attrib3="2" attrib4="1">Even more text</DifferentElement>
% </XMLname>
%
% Will produce:
% s.XMLname.Attributes.attrib1 = "Some value";
% s.XMLname.Element.Text = "Some text";
% s.XMLname.DifferentElement{1}.Attributes.attrib2 = "2";
% s.XMLname.DifferentElement{1}.Text = "Some more text";
% s.XMLname.DifferentElement{2}.Attributes.attrib3 = "2";
% s.XMLname.DifferentElement{2}.Attributes.attrib4 = "1";
% s.XMLname.DifferentElement{2}.Text = "Even more text";
%
% Please note that the following characters are substituted
% '-' by '_dash_', ':' by '_colon_' and '.' by '_dot_'
%
% Written by W. Falkena, ASTI, TUDelft, 21-08-2010
% Attribute parsing speed increased by 40% by A. Wanner, 14-6-2011
% Added CDATA support by I. Smirnov, 20-3-2012
%
% Modified by X. Mo, University of Wisconsin, 12-5-2012
if (nargin < 1)
clc;
help xml2struct
return
end
if isa(file, 'org.apache.xerces.dom.DeferredDocumentImpl') || isa(file, 'org.apache.xerces.dom.DeferredElementImpl')
% input is a java xml object
xDoc = file;
else
%check for existance
if (exist(file,'file') == 0)
%Perhaps the xml extension was omitted from the file name. Add the
%extension and try again.
if (isempty(strfind(file,'.xml')))
file = [file '.xml'];
end
if (exist(file,'file') == 0)
error(['The file ' file ' could not be found']);
end
end
%read the xml file
xDoc = xmlread(file);
end
%parse xDoc into a MATLAB structure
s = parseChildNodes(xDoc);
end
% ----- Subfunction parseChildNodes -----
function [children,ptext,textflag] = parseChildNodes(theNode)
% Recurse over node children.
children = struct;
ptext = struct; textflag = 'Text';
if hasChildNodes(theNode)
childNodes = getChildNodes(theNode);
numChildNodes = getLength(childNodes);
for count = 1:numChildNodes
theChild = item(childNodes,count-1);
[text,name,attr,childs,textflag] = getNodeData(theChild);
if (~strcmp(name,'#text') && ~strcmp(name,'#comment') && ~strcmp(name,'#cdata_dash_section'))
%XML allows the same elements to be defined multiple times,
%put each in a different cell
if (isfield(children,name))
if (~iscell(children.(name)))
%put existsing element into cell format
children.(name) = {children.(name)};
end
index = length(children.(name))+1;
%add new element
children.(name){index} = childs;
if(~isempty(fieldnames(text)))
children.(name){index} = text;
end
if(~isempty(attr))
children.(name){index}.('Attributes') = attr;
end
else
%add previously unknown (new) element to the structure
children.(name) = childs;
if(~isempty(text) && ~isempty(fieldnames(text)))
children.(name) = text;
end
if(~isempty(attr))
children.(name).('Attributes') = attr;
end
end
else
ptextflag = 'Text';
if (strcmp(name, '#cdata_dash_section'))
ptextflag = 'CDATA';
elseif (strcmp(name, '#comment'))
ptextflag = 'Comment';
end
%this is the text in an element (i.e., the parentNode)
if (~isempty(regexprep(text.(textflag),'[\s]*','')))
if (~isfield(ptext,ptextflag) || isempty(ptext.(ptextflag)))
ptext.(ptextflag) = text.(textflag);
else
%what to do when element data is as follows:
%<element>Text <!--Comment--> More text</element>
%put the text in different cells:
% if (~iscell(ptext)) ptext = {ptext}; end
% ptext{length(ptext)+1} = text;
%just append the text
ptext.(ptextflag) = [ptext.(ptextflag) text.(textflag)];
end
end
end
end
end
end
% ----- Subfunction getNodeData -----
function [text,name,attr,childs,textflag] = getNodeData(theNode)
% Create structure of node info.
%make sure name is allowed as structure name
name = toCharArray(getNodeName(theNode))';
name = strrep(name, '-', '_dash_');
name = strrep(name, ':', '_colon_');
name = strrep(name, '.', '_dot_');
attr = parseAttributes(theNode);
if (isempty(fieldnames(attr)))
attr = [];
end
%parse child nodes
[childs,text,textflag] = parseChildNodes(theNode);
if (isempty(fieldnames(childs)) && isempty(fieldnames(text)))
%get the data of any childless nodes
% faster than if any(strcmp(methods(theNode), 'getData'))
% no need to try-catch (?)
% faster than text = char(getData(theNode));
text.(textflag) = toCharArray(getTextContent(theNode))';
end
end
% ----- Subfunction parseAttributes -----
function attributes = parseAttributes(theNode)
% Create attributes structure.
attributes = struct;
if hasAttributes(theNode)
theAttributes = getAttributes(theNode);
numAttributes = getLength(theAttributes);
for count = 1:numAttributes
%attrib = item(theAttributes,count-1);
%attr_name = regexprep(char(getName(attrib)),'[-:.]','_');
%attributes.(attr_name) = char(getValue(attrib));
%Suggestion of Adrian Wanner
str = toCharArray(toString(item(theAttributes,count-1)))';
k = strfind(str,'=');
attr_name = str(1:(k(1)-1));
attr_name = strrep(attr_name, '-', '_dash_');
attr_name = strrep(attr_name, ':', '_colon_');
attr_name = strrep(attr_name, '.', '_dot_');
attributes.(attr_name) = str((k(1)+2):(end-1));
end
end
end