I was wondering what the best way to sort a chart using the column command to center each column instead of the default left aligned column was. I have been using the column -t filename command.
Current Output:
Label1 label2
Anotherlabel label2442
label152 label42242
label78765 label373737737
Desired Output: Something like this
Label1 label2
Anotherlabel label2442
label152 label42242
label78765 label373737737
Basically, I want it to be centered instead of left aligned.
Here is an awk solution:
# Collect all lines in "data", keep track of maximum width for each field
{
data[NR] = $0
for (i = 1; i <= NF; ++i)
max[i] = length($i) > max[i] ? length($i) : max[i]
}
END {
for (i = 1; i <= NR; ++i) {
# Split record into array "arr"
split(data[i], arr)
# Loop over array
for (j = 1; j <= NF; ++j) {
# Calculate amount of padding required
pad = max[j] - length(arr[j])
# Print field with appropriate padding, see below
printf "%*s%*s%s", length(arr[j]) + int(pad/2), arr[j], \
pad % 2 == 0 ? pad/2 : int(pad/2) + 1, "", \
j == NF ? "" : " "
}
# Newline at end of record
print ""
}
}
Called like this:
$ awk -f centre.awk infile
Label1 label2
Anotherlabel label2442
label152 label42242
label78765 label373737737
The printf statement uses padding with dynamic widths:
The first %*s takes care of left padding and the data itself: arr[j] gets printed and padded to a total width of length(arr[j]) + int(pad/2).
The second %*s prints the empty string, left padded to half of the total padding required. pad % 2 == 0 ? pad/2 : int(pad/2) + 1 checks if the total padding was an even number, and if not, adds an extra space.
The last %s prints j == NF ? "" : " ", i.e., two spaces, unless we're at the last field.
Some older awks don't support the %*s syntax, but the formatting string can be assembled like width = 5; "%" width "s" in that case.
Here's a Python program to do what you want. It's probably too hard to do in bash, so you'll need to use a custom program or awk script. Basic algorithm:
count number of columns
[optional] make sure each line has the same number of columns
figure out the maximum length of data for each column
print each line using the max lengths
.
#!/usr/bin/env python3
import sys
def column():
# Read file and split each line into fields (by whitespace)
with open(sys.argv[1]) as f:
lines = [line.split() for line in f]
# Check that each line has the same number of fields
num_fields = len(lines[0])
for n, line in enumerate(lines):
if len(line) != num_fields:
print('Line {} has wrong number of columns: expected {}, got {}'.format(n, num_fields, len(line)))
sys.exit(1)
# Calculate the maximum length of each field
max_column_widths = [0] * num_fields
for line in lines:
line_widths = (len(field) for field in line)
max_column_widths = [max(z) for z in zip(max_column_widths, line_widths)]
# Now print them centered using the max_column_widths
spacing = 4
format_spec = (' ' * spacing).join('{:^' + str(n) + '}' for n in max_column_widths)
for line in lines:
print(format_spec.format(*line))
if __name__ == '__main__':
column()
I have a binary file of an image which is 256x256 array
I just want to add a space after every byte in this file
e.g 000101010101110101010100
expected output: 00010101 01011101 01010100
can any one suggest me to this using matlab prefarably
If you already have those 0/1 values in the form of a char array, such as
a =
101001011111011101010100
100101001010100000111111
100100100001011110110101
001010011111110000001010
you can do it this way:
n = 8; %// byte size
b = a(:,ceil(n/(n+1):n/(n+1):end)); %// repeat every n-th position
b(:,n+1:n+1:end) = ' '; %// replace those repeated positions by spaces
which gives
b =
10100101 11110111 01010100
10010100 10101000 00111111
10010010 00010111 10110101
00101001 11111100 00001010
... and to get the char array mentioned by Luis Mendo you can do the following:
%edge length of image
edge = 256;
%open and read file
fid = fopen('imageData.txt');
C = fread(fid,[edge+2,edge]);
The +2 actually depends on your file, does it contain carriage return? If not, leave it out.
% get char array
chararray = char(C(1:edge,:)');
Then I would proceed as follows:
wordlength = 8;
cellarray = mat2cell(chararray,ones(1,edge),wordlength*ones(1,edge/wordlength));
cellarray = cellfun(#(x) [x ' '],cellarray,'uni',0);
output = cell2mat(cellarray);
%// remove last column with unwanted space
output = output(:,end-1:end)
and write that to a file again.
regarding your comment, try this:
wordlength = 8;
cellarray = mat2cell(chararray,ones(1,edge),wordlength*ones(1,edge/wordlength));
fid2 = fopen( yourFilepath ,'w');
fprintf(fid2 , '%s\r\n', cellarray {:});
fclose(fid2);
You can do it with a simple Perl one-liner:
perl -e '$/=\1;while(<>){print $_," ";}' < yourFile > newfile
The $/=\1 sets the record separator so we read one byte at a time. Then we go into a loop reading one byte at a time, printing it out followed by a space till end of file.
Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 9 years ago.
Improve this question
No idea if this is ok to ask here since it's not programming but I have no idea where else to go:
I want to organise the following data in a consistent way. At the moment it's a mess, with only the first two columns (comma separated) consistent. The remaining columns can number anywhere from 1-9 and are usually different.
In other words, I want to sort it so the text matches (all of the value columns in a row, all of the recoil columns in a row, etc). Then I can remove the text and add a header, and it will still make sense.
bm_wp_upg_o_t1micro, sight, value = 3, zoom = 3, recoil = 1, spread_moving = -1
bm_wp_upg_o_marksmansight_rear, sight, value = 3, zoom = 1, recoil = 1, spread = 1
bm_wp_upg_o_marksmansight_front, extra, value = 1
bm_wp_m4_upper_reciever_edge, upper_reciever, value = 3, recoil = 1
bm_wp_m4_upper_reciever_round, upper_reciever, value = 1
bm_wp_m4_uupg_b_long, barrel, value = 4, damage = 1, spread = 1, spread_moving = -2, concealment = -2
Any suggestions (even on just where the right place is to actually ask this) would be great.
Context is just raw data ripped from a game file that I'm trying to organise.
I'm afraid regex isn't going to help you much here because of the irregular nature of your input (it would be possible to match it, but it would be a bear to get it all arranged one way or another). This could be done pretty easily with any programming language, but for stuff like this, I always go to awk.
Assuming your input is in a file called input.txt, put the following in a program called parse.awk:
BEGIN {
FS=" *, *";
formatStr = "%32s,%8s,%8s,%8s,%10s,%16s,%8s,%18s,%10s,%10s,%16s,%16s\n";
printf( formatStr, "id", "sight", "value", "zoom", "recoil", "spread_moving", "extra", "upper_receiver", "barrel", "damage", "spread_moving", "concealment" );
}
{
split("",a);
for( i=2; i<=NF; i++ ) {
if( split( $(i), kvp, " *= *" ) == 1 ) {
a[kvp[1]] = "x";
} else {
a[kvp[1]] = gensub( /^\s*|\s*$/, "", "g", kvp[2] );
}
}
printf( formatStr, $1, a["sight"], a["value"], a["zoom"], a["recoil"],
a["spread_moving"], a["extra"], a["upper_receiver"],
a["barrel"], a["damage"], a["spread_moving"], a["concealment"] );
}
Run awk against it:
awk -f parse.awk input.txt
And get your output:
id, sight, value, zoom, recoil, spread_moving, extra, upper_receiver, barrel, damage, spread_moving, concealment
bm_wp_upg_o_t1micro, x, 3, 3, 1, -1, , , , , -1,
bm_wp_upg_o_marksmansight_rear, x, 3, 1, 1, , , , , , ,
bm_wp_upg_o_marksmansight_front, , 1, , , , x, , , , ,
bm_wp_m4_upper_reciever_edge, , 3, , 1, , , , , , ,
bm_wp_m4_upper_reciever_round, , 1, , , , , , , , ,
bm_wp_m4_uupg_b_long, , 4, , , -2, , , x, 1, -2, -2
Note that I chose to just use an 'x' for sight, which seems to a present/absent thing. You can use whatever you want there.
If you're using Linux or a Macintosh, you should have awk available. If you're on Windows, you'll have to install it.
I did make another awk version. I think this should a little easier to read.
All value/column are read from the file to make it as dynamic as possible.
awk -F, '
{
ID[$1]=$2 # use column 1 as index
for (i=3;i<=NF;i++ ) # loop through all fields from #3 to end
{
gsub(/ +/,"",$i) # remove space from field
split($i,a,"=") # split field in name and value a[1] and a[2]
COLUMN[a[1]]++ # store field name as column name
DATA[$1" "a[1]]=a[2] # store data value in DATA using field #1 and column name as index
}
}
END {
printf "%49s ","info" # print info
for (i in COLUMN)
{printf "%15s",i} # print column name
print ""
for (i in ID) # loop through all ID
{
printf "%32s %16s ",i, ID[i] # print ID and info
for (j in COLUMN)
{
printf "%14s ",DATA[i" "j]+0 # print value
}
print ""
}
}' file
Output
info spread recoil zoom concealment spread_moving damage value
bm_wp_m4_upper_reciever_round upper_reciever 0 0 0 0 0 0 1
bm_wp_m4_uupg_b_long barrel 1 0 0 -2 -2 1 4
bm_wp_upg_o_marksmansight_rear sight 1 1 1 0 0 0 3
bm_wp_upg_o_marksmansight_front extra 0 0 0 0 0 0 1
bm_wp_m4_upper_reciever_edge upper_reciever 0 1 0 0 0 0 3
bm_wp_upg_o_t1micro sight 0 1 3 0 -1 0 3
Stick with Ethan's answer — this is just me enjoying myself. (And yes, that makes me pretty weird!)
awk script
awk 'BEGIN {
# f_idx[field] holds the column number c for a field=value item
# f_name[c] holds the names
# f_width[c] holds the width of the widest value (or the field name)
# f_fmt[c] holds the appropriate format
FS = " *, *"; n = 2;
f_name[0] = "id"; f_width[0] = length(f_name[0])
f_name[1] = "type"; f_width[1] = length(f_name[1])
}
{
#-#print NR ":" $0
line[NR,0] = $1
len = length($1)
if (len > f_width[0])
f_width[0] = len
line[NR,1] = $2
len = length($2)
if (len > f_width[1])
f_width[1] = len
for (i = 3; i <= NF; i++)
{
split($i, fv, " = ")
#-#print "1:" fv[1] ", 2:" fv[2]
if (!(fv[1] in f_idx))
{
f_idx[fv[1]] = n
f_width[n++] = length(fv[1])
}
c = f_idx[fv[1]]
f_name[c] = fv[1]
gsub(/ /, "", fv[2])
len = length(fv[2])
if (len > f_width[c])
f_width[c] = len
line[NR,c] = fv[2]
#-#print c ":" f_name[c] ":" f_width[c] ":" line[NR,c]
}
}
END {
for (i = 0; i < n; i++)
f_fmt[i] = "%s%" f_width[i] "s"
#-#for (i = 0; i < n; i++)
#-# printf "%d: (%d) %s %s\n", i, f_width[i], f_name[i], f_fmt[i]
#-# pad = ""
for (j = 0; j < n; j++)
{
printf f_fmt[j], pad, f_name[j]
pad = ","
}
printf "\n"
for (i = 1; i <= NR; i++)
{
pad = ""
for (j = 0; j < n; j++)
{
printf f_fmt[j], pad, line[i,j]
pad = ","
}
printf "\n"
}
}' data
This script adapts to the data it finds in the file. It assigns the column heading 'id' to column 1 of the input, and 'type' to column 2. For each of the sets of values in columns 3..N, it splits up the data into key (in fv[1]) and value (in fv[2]). If the key has not been seen before, it is assigned a new column number, and the key is stored as the column name, and the width of key as the initial column width. Then the value is stored in the appropriate column within the line.
When all the data's read, the script knows what the column headings are going to be. It can then create a set of format strings. Then it prints the headings and all the rows of data. If you don't want fixed width output, then you can simplify the script considerably. There are some (mostly minor) simplifications that could be made to this script.
Data file
bm_wp_upg_o_t1micro, sight, value = 3, zoom = 3, recoil = 1, spread_moving = -1
bm_wp_upg_o_marksmansight_rear, sight, value = 3, zoom = 1, recoil = 1, spread = 1
bm_wp_upg_o_marksmansight_front, extra, value = 1
bm_wp_m4_upper_receiver_edge, upper_receiver, value = 3, recoil = 1
bm_wp_m4_upper_receiver_round, upper_receiver, value = 1
bm_wp_m4_uupg_b_long, barrel, value = 4, damage = 1, spread = 1, spread_moving = -2, concealment = -2
Output
id, type,value,zoom,recoil,spread_moving,spread,damage,concealment
bm_wp_upg_o_t1micro, sight, 3, 3, 1, -1, , ,
bm_wp_upg_o_marksmansight_rear, sight, 3, 1, 1, , 1, ,
bm_wp_upg_o_marksmansight_front, extra, 1, , , , , ,
bm_wp_m4_upper_receiver_edge,upper_receiver, 3, , 1, , , ,
bm_wp_m4_upper_receiver_round,upper_receiver, 1, , , , , ,
bm_wp_m4_uupg_b_long, barrel, 4, , , -2, 1, 1, -2
I am using Excel 2007 which supports Columns upto 16,384 Columns. I would like to obtain the Column name corresponding Column Number.
Currently, I am using the following code. However this code supports upto 256 Columns. Any idea how to obtain Column Name if the column number is greater than 256.
function loc = xlcolumn(column)
if isnumeric(column)
if column>256
error('Excel is limited to 256 columns! Enter an integer number <256');
end
letters = {'A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z'};
count = 0;
if column-26<=0
loc = char(letters(column));
else
while column-26>0
count = count + 1;
column = column - 26;
end
loc = [char(letters(count)) char(letters(column))];
end
else
letters = ['A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z'];
if size(column,2)==1
loc =findstr(column,letters);
elseif size(column,2)==2
loc1 =findstr(column(1),letters);
loc2 =findstr(column(2),letters);
loc = (26 + 26*loc1)-(26-loc2);
end
end
Thanks
As a diversion, here is an all function handle example, with (almost) no file-based functions required. This is based on the dec2base function, since Excel column names are (almost) base 26 numbers, with the frustrating difference that there are no "0" characters.
Note: this is probably a terrible idea overall, but it works. Better solutions are probably found elsewhere in the file exchange.
First, the one file based function that I couldn't get around, to perform arbitrary depth function composition.
function result = compose( fnHandles )
%COMPOSE Compose a set of functions
% COMPOSE({fnHandles}) returns a function handle consisting of the
% composition of the cell array of input function handles.
%
% For example, if F, G, and H are function handles with one input and
% one output, then:
% FNCOMPOSED = COMPOSE({F,G,H});
% y = FNCOMPOSED(x);
% is equivalent to
% y = F(G(H(x)));
if isempty(fnHandles)
result = #(x)x;
elseif length(fnHandles)==1
result = fnHandles{1};
else
fnOuter = fnHandles{1};
fnRemainder = compose(fnHandles(2:end));
result = #(x)fnOuter(fnRemainder(x));
end
Then, the bizarre, contrived path to convert base26 values into the correct string
%Functions leading to "getNumeric", which creates a numeric, base26 array
remapUpper = #(rawBase)(rawBase + (rawBase>='A')*(-55)); %Map the letters 'A-P' to [10:26]
reMapLower = #(rawBase)(rawBase + (rawBase<'A')*(-48)); %Map characters '0123456789' to [0:9]
getRawBase = #(x)dec2base(x, 26);
getNumeric = #(x)remapUpper(reMapLower(getRawBase(x)));
%Functions leading to "correctNumeric"
% This replaces zeros with 26, and reduces the high values entry by 1.
% Similar to "borrowing" as we learned in longhand subtraction
borrowDownFrom = #(x, fromIndex) [x(1:(fromIndex-1)) (x(fromIndex)-1) (x(fromIndex+1)+26) (x((fromIndex+2):end))];
borrowToIfNeeded = #(x, toIndex) (x(toIndex)<=0)*borrowDownFrom(x,toIndex-1) + (x(toIndex)>0)*(x); %Ugly numeric switch
getAllConditionalBorrowFunctions = #(numeric)arrayfun(#(index)#(numeric)borrowToIfNeeded(numeric, index),(2:length(numeric)),'uniformoutput',false);
getComposedBorrowFunction = #(x)compose(getAllConditionalBorrowFunctions(x));
correctNumeric = #(x)feval(getComposedBorrowFunction(x),x);
%Function to replace numerics with letters, and remove leading '#' (leading
%zeros)
numeric2alpha = #(x)regexprep(char(x+'A'-1),'^#','');
%Compose complete function
num2ExcelName = #(x)arrayfun(#(x)numeric2alpha(correctNumeric(getNumeric(x))), x, 'uniformoutput',false)';
Now test using some stressing transitions:
>> num2ExcelName([1:5 23:28 700:704 727:729 1024:1026 1351:1355 16382:16384])
ans =
'A'
'B'
'C'
'D'
'E'
'W'
'X'
'Y'
'Z'
'AA'
'AB'
'ZX'
'ZY'
'ZZ'
'AAA'
'AAB'
'AAY'
'AAZ'
'ABA'
'AMJ'
'AMK'
'AML'
'AYY'
'AYZ'
'AZA'
'AZB'
'AZC'
'XFB'
'XFC'
'XFD'
This function I wrote works for any number of columns (until Excel runs out of columns). It just requires a column number input (e.g. 16368 will return a string 'XEN').
If the application of this concept is different than my function, it's important to note that a column of x number of A's begins every 26^(x-1) + 26^(x-2) + ... + 26^2 + 26 + 1. (e.g. 'AAA' begins on 26^2 + 26 + 1 = 703)
function [col_str] = let_loc(num_loc)
test = 2;
old = 0;
x = 0;
while test >= 1
old = 26^x + old;
test = num_loc/old;
x = x + 1;
end
num_letters = x - 1;
str_array = zeros(1,num_letters);
for i = 1:num_letters
loc = floor(num_loc/(26^(num_letters-i)));
num_loc = num_loc - (loc*26^(num_letters-i));
str_array(i) = char(65 + (loc - 1));
end
col_str = strcat(str_array(1:length(str_array)));
end
Hope this saves someone some time!