get a mean of indexed data without using for loop
in [Matlab]
|
From: Ryan on 20 May 2010 09:14 I'm trying to obtain a mean of numbers, using different indices, without using a 'for' loop. For example, say I have the following matrix x: x = [101 3; ... 101 2; ... 101 1; ... 102 8; ... 103 5; ... 103 5; ... 104 6; ... 104 8; .. 104 5.2; ... 104 7; ... 109 5; ... 109 7]; I would like to get the mean in column 2 for all instances of column 1==101, and the mean in column 2 for all instances of column 1==101, etc. Does anyone know of an efficient manner for doing this without using a 'for' loop?
From: mat001 on 20 May 2010 09:38 "Ryan " <jessupr(a)tcd.ie> wrote in message <ht3cis$r6u$1(a)fred.mathworks.com>... > I'm trying to obtain a mean of numbers, using different indices, without using a 'for' loop. For example, say I have the following matrix x: > x = [101 3; ... > 101 2; ... > 101 1; ... > 102 8; ... > 103 5; ... > 103 5; ... > 104 6; ... > 104 8; .. > 104 5.2; ... > 104 7; ... > 109 5; ... > 109 7]; > > I would like to get the mean in column 2 for all instances of column 1==101, and the mean in column 2 for all instances of column 1==101, etc. Does anyone know of an efficient manner for doing this without using a 'for' loop? Try mean(x(:,:)) %may be this is what you asked for.
From: Sean on 20 May 2010 09:42 "Ryan " <jessupr(a)tcd.ie> wrote in message <ht3cis$r6u$1(a)fred.mathworks.com>... > I'm trying to obtain a mean of numbers, using different indices, without using a 'for' loop. For example, say I have the following matrix x: > x = [101 3; ... > 101 2; ... > 101 1; ... > 102 8; ... > 103 5; ... > 103 5; ... > 104 6; ... > 104 8; .. > 104 5.2; ... > 104 7; ... > 109 5; ... > 109 7]; > > I would like to get the mean in column 2 for all instances of column 1==101, and the mean in column 2 for all instances of column 1==101, etc. Does anyone know of an efficient manner for doing this without using a 'for' loop? One way: x = [101 3; ... 101 2; ... 101 1; ... 102 8; ... 103 5; ... 103 5; ... 104 6; ... 104 8; ... 104 5.2; ... 104 7; ... 109 5; ... 109 7]; ndxs = unique(x(:,1)); bins = histc(x(:,1),ndxs); C = mat2cell(x(:,2),bins); my_means = cellfun(@mean,C);
From: Ryan on 21 May 2010 07:30 "Sean " <sean.dewolski(a)nospamplease.umit.maine.edu> wrote in message <ht3e7c$gjv$1(a)fred.mathworks.com>... > > > One way: > > x = [101 3; ... > 101 2; ... > 101 1; ... > 102 8; ... > 103 5; ... > 103 5; ... > 104 6; ... > 104 8; ... > 104 5.2; ... > 104 7; ... > 109 5; ... > 109 7]; > > ndxs = unique(x(:,1)); > bins = histc(x(:,1),ndxs); > C = mat2cell(x(:,2),bins); > > my_means = cellfun(@mean,C); Thanks, Sean, that works perfectly! I've had need for this many times so it will come in very handy.
From: Jos (10584) on 21 May 2010 11:25 "Ryan " <jessupr(a)tcd.ie> wrote in message <ht5qse$t4v$1(a)fred.mathworks.com>... > "Sean " <sean.dewolski(a)nospamplease.umit.maine.edu> wrote in message <ht3e7c$gjv$1(a)fred.mathworks.com>... > > > > > > > One way: > > > > x = [101 3; ... > > 101 2; ... > > 101 1; ... > > 102 8; ... > > 103 5; ... > > 103 5; ... > > 104 6; ... > > 104 8; ... > > 104 5.2; ... > > 104 7; ... > > 109 5; ... > > 109 7]; > > > > ndxs = unique(x(:,1)); > > bins = histc(x(:,1),ndxs); > > C = mat2cell(x(:,2),bins); > > > > my_means = cellfun(@mean,C); > > Thanks, Sean, that works perfectly! I've had need for this many times so it will come in very handy. Here is another, faster way using ACCUMARRAY directly instead of CELLFUN: [ix,ix,ix] = unique(x(:,1)) ; % equals [unused1, unused2,ix] = unique(x(:,1)) C = accumarray(ix,x(:,2),[],@mean) hth Jos
|
Pages: 1 Prev: GUI slider for displaying images Next: steps to convert RGB (colored image) to binary? |