I have an N-dimensional array of items whose last dimension is the index of the array.
For example, if the array A contained images, then A(:,:,:,1) would be the first image, A(:,:,:,2) would be the second image, and so forth.
Similarly, if the array just contained integers, then A(:,1) would be the first integer, A(:,2) would be the second integer, and so forth.
-=-=-=-
What I'm trying to do is delete the first item from A when I do not know ahead of time what dimensionality it is.
If A contains images, I want to do this:
A(:,:,:,1) = [];
If A contains integers, I want to do this:
A(:,1) = [];
The problem is since I don't know what dimensionality it is, I don't know how many colons to put, and I don't know how to denote "N-1 colons here" in Matlab.
I'm hoping there is a programmatic way to do this, but I frankly have no idea what to search for if this is possible.
You can either use cell to comma-separated list expansion:
%// Build cell: {':', ':', ..., ':', [1]}
I(1:ndims(A)-1) = {':'};
I{ndims(A)} = 1;
%// Expand cell to comma separated list and delete:
A(I{:}) = [];
Or convert to cell using num2cell and then convert back using cell2mat:
C = num2cell(A,1:ndims(A)-1);
A = cell2mat(C(2:end));
I guess that unless you really need n-dimensional arrays, doing this with a cell array of n-1 dimensional arrays instead (as is C in the above code) should be a smart move in terms of simplicity of notation.
Related
A fourier analysis I'm doing outputs 5 data fields, each of which I've collected into 1-d numpy arrays: freq bin #, amplitude, wavelength, normalized amplitude, %power.
How best to structure the data so I can sort by descending amplitude?
When testing with just one data field, I was able to use a dict as follows:
fourier_tuples = zip(range(len(fourier)), fourier)
fourier_map = dict(fourier_tuples)
import operator
fourier_sorted = sorted(fourier_map.items(), key=operator.itemgetter(1))
fourier_sorted = np.argsort(-fourier)[:3]
My intent was to add the other arrays to line 1, but this doesn't work since dicts only accept 2 terms. (That's why this post doesn't solve my issue.)
Stepping back, is this a reasonable approach, or are there better ways to combine & sort separate arrays? Ultimately, I want to take the data values from the top 3 freqs and associated other data, and write them to an output data file.
Here's a snippet of my data:
fourier = np.array([1.77635684e-14, 4.49872050e+01, 1.05094837e+01, 8.24322470e+00, 2.36715913e+01])
freqs = np.array([0. , 0.00246951, 0.00493902, 0.00740854, 0.00987805])
wavelengths = np.array([inf, 404.93827165, 202.46913583, 134.97942388, 101.23456791])
amps = np.array([4.33257766e-16, 1.09724890e+00, 2.56328871e-01, 2.01054261e-01, 5.77355886e-01])
powers% = np.array([4.8508237956526163e-32, 0.31112370227749603, 0.016979224022185751, 0.010445983875848858, 0.086141014686372669])
The last 4 arrays are other fields corresponding to 'fourier'. (Actual array lengths are 42, but pared down to 5 for simplicity.)
You appear to be using numpy, so here is the numpy way of doing this. You have the right function np.argsort in your post, but you don't seem to use it correctly:
order = np.argsort(amplitudes)
This is similar to your dictionary trick only it computes the inverse shuffling compared to your procedure. Btw. why go through a dictionary and not simply a list of tuples?
The contents of order are now indices into amplitudes the first cell of order contains the position of the smallest element of amplitudes, the second cell contains the position of the next etc. Therefore
top5 = order[:-6:-1]
Provided your data are 1d numpy arrays you can use top5 to extract the elements corresponding to the top 5 ampltiudes by using advanced indexing
freq_bin[top5]
amplitudes[top5]
wavelength[top5]
If you want you can group them together in columns and apply top5 to the resulting n-by-5 array:
np.c_[freq_bin, amplitudes, wavelength, ...][top5, :]
If I understand correctly you have 5 separate lists of the same length and you are trying to sort all of them based on one of them. To do that you can either use numpy or do it with vanilla python. Here are two examples from top of my head (sorting is based on the 2nd list).
a = [11,13,10,14,15]
b = [2,4,1,0,3]
c = [22,20,23,25,24]
#numpy solution
import numpy as np
my_array = np.array([a,b,c])
my_sorted_array = my_array[:,my_array[1,:].argsort()]
#vanilla python solution
from operator import itemgetter
my_list = zip(a,b,c)
my_sorted_list = sorted(my_list,key=itemgetter(1))
You can then flip the array with my_sorted_array = np.fliplr(my_sorted_array) if you wish or if you are working with lists you can reverse it in place with my_sorted_list.reverse()
EDIT:
To get first n values only, you have to simply slice the array similarly to what #Paul is suggesting. Slice is done in a similar manner to classic list slicing by specifying start:stop:step (you can omit the step) arguments. In your case for 5 top columns it would be [:,-5:]. So in the example above you can take top 2 columns from each row like this:
my_sliced_sorted_array = my_sorted_array[:,-2:]
result will be:
array([[15, 13],
[ 3, 4],
[24, 20]])
Hope it helps.
I am trying to carry out the intersection of two arrays in Matlab but I cannot find the way.
The arrays that I want to intersect are:
and
I have tried:[dur, itimes, inewtimes ] = intersect(array2,char(array1));
but no luck.
However, if I try to intersect array1 with array3 (see array3 below), [dur, itimes, inewtimes ] = intersect(array3,char(array1));the intersection is performed without any error.
Why I cannot intersect array1 with array2?, how could I do it?. Thank you.
Just for ease of reading, your formats for Arrays are different, and you want to make them the same. There are many options for you, like #Visser suggested, you could convert the date/time into a long int which allows faster computation, or you can keep them as strings, or even convert them into characters (like what you have done with char(Array2)).
This is my example:
A = {'00:00:00';'00:01:01'} %//Type is Cell String
Z = ['00:00:00';'00:01:01'] %//Type is Cell Char
Q = {{'00:00:00'};{'00:01:01'}} %//Type is a Cell of Cells
A = cellstr(A) %//Convert CellStr to CellStr is essentially doing nothing
Z = cellstr(Z) %//Convert CellChar to CellStr
Q = vertcat(Q{:,:}) %// Convert Cell of Cells to Cell of Strings
I = intersect (A,Z)
>>'00:00:00'
'00:01:01'
II = intersect (A,Q)
>>'00:00:00'
'00:01:01'
This keeps your dates in the format of Strings in case you want to export them back into a txt/csv file.
Your first array would look something like this:
array1 = linspace(0,1,86400); % creates 86400 seconds in 1 day
Your second array should be converted using datenum, then use cell2mat to make it a matrix. Lastly, use ismember to find the intersection:
InterSect = ismember(array2,array1);
I have two cell arrays, the sizes are 1x20033 and 1x19. Let's call these two cell arrays as A and B. I want to compare each cell of A with each cell of B to see if there is any common element.
Finally, I need to build a binary matrix and put one when there is a match.
I tried this:
BinaryMatrix=zeros(20033,19);
for i=1:1:20033
for j=1:1:19
match=find(ismember(A{i},B{j}));
if match==1
BinaryMatrix(i,j)= 1;
end
end
end
but I faced this error: "Input A of class double and input B of class cell must be
cell arrays of strings, unless one is a string."
Please tell me What should I do to solve it?
The code that you have almost works. What I would recommend you do is split up the strings found in A and B by spaces. As such, A and B would then be cell arrays of elements where each element in A or B is a single word. The spaces will serve as delimiters for separating out the words.
Once you do this, use intersect to see if there are any common words between the words in A and the words in B. intersect works by considering two arrays (these can be numeric arrays, cell arrays, etc.) C and D as sets, and it returns the set intersection between these two arrays.
In our case, C and D would be a cell array of words separated by spaces from A and B. intersect(C,D) will return a cell array of strings where each element in the output is a string found in both C and D. As such, should this cell array be non-empty, we have found at least one common word between C and D. If this is the case, then set your binary flag at the location of your matrix to 1. In other words:
BinaryMatrix = false(20033,19);
for i=1:1:20033
for j=1:1:19
Asplit = strsplit(A{i});
Bsplit = strsplit(B{j});
if (~isempty(intersect(Asplit, Bsplit)))
BinaryMatrix(i,j)= true;
end
end
end
You'll notice that I have changed your matrix from zeros(20033,19), to false(20033,19). The reason why is because by doing zeros, you are allocating 8 bytes per number in your matrix as this will create your matrix in double precision. By doing false, this will be a logical matrix instead, and you are allocating 1 byte per number. Seeing as how you want BinaryMatrix to be either true or false, don't use double - use logical. I don't know how large both cell arrays are, and so doing this will cut down your memory consumption by 8.
Minor Note
strsplit is only available from R2013a and onwards. If you have a version of MATLAB that is R2012b and lower, replace strsplit with regexp. As such, you would replace the two lines in the for loop with:
Asplit = regexp(A{i}, ' ', 'split');
Bsplit = regexp(B{j}, ' ', 'split');
Going nuts with cell array, because I just can't get rid of it... However, it will be an easy one for you guys out here.
So here is why:
I have a dataset (data) which contains two variables: A (Numbers) and B (cell array).
Unfortunately I can't even reconstruct the problem nevertheless my imported table looks like this:
data=dataset;
data.A = [1;1;3;3;3];
data.B = ['A';'A';'BUU';'BUU';'A'];
where data.B is of the type 5x1 cell which I can't reconstruct
all I want now is the unique rows like
ans= [1 A;3 BUU;3 A]
the result should be in a dataset or just two vectors where the rows are equivalent.
but unique([dataA dataB],'rows') can't handle cell arrays and I can't find anywhere in the www how I simple convert the cell array B to a vector of strings (does it exist?).
cell2mat() didn't work for me, because of the different word length ('A' vs 'BUU').
Though, two things I would love to learn: Making an 5x1 cell to an string vector
and find unique rows out of numbers and strings (or cells).
Thank you very much!
Cheers Dominik
The problem is that the A and B fields are of a different type. Although they could be concatenated into a cell array, unique can't handle that. A general trick for cases like this is to "translate" elements of each field (column) to unique identifiers, i.e. numbers. This translation can be done applying unique to each field separately and getting its third output. The obtained identifiers can now be concatenated into a matrix, so that each row of this matrix is a "composite identifier". Finally, unique with 'rows' option can be applied to this matrix.
So, in your case:
[~, ~, kA] = unique(data.A);
[~, ~, kB] = unique(data.B);
[~, jR] = unique([kA kB], 'rows');
Now build the result as (same format as data)
result.A = data.A(jR);
result.B = data.B(jR);
or as (2D cell array)
result = cat(2, mat2cell(data.A(jR), ones(1,numel(jR))), data.B(jR));
Here is my clumpsy solution
tt.A = [1;1;3;3;3];
tt.B = {'A';'A';'BUU';'BUU';'A'};
Convert integers to characters, then merge and find unique strings
tt.C = cellstr(num2str(tt.A));
tt.D = cellfun(#(x,y) [x y],tt.C,tt.B,'UniformOutput',0);
[tt.F,tt.E] = unique(tt.D);
Display results
tt.F
I have a cell matrix with 2 rows - sorted cell array of different strings and array of numbers. Also I have an example string. It's guarantied that this string appears in the 1st row in the cell array. I want to get index of example's appearance in cell array of strings.
Is there any function in Matlab, that provides solving with logarithmic complexity (something like binary search)?
If you look at ismember code (type open ismember), you'll see that it basically
Checks if the array is sorted (by calling issorted);
If not, it sorts the array;
Then it applies binary search.
So you can directly use ismember.
Example:
>> strings = {'a', 'aa', 'be', 'day', 'yes'};
>> [tf, loc ] = ismember('day', strings);
>> loc
loc =
4
Or maybe modify ismember (save it with another name) to bypass step 1, since you already know your array is sorted.