I have files with hundreds of sequences with frequency values reported as "Freq X" and missing characters represented by a dash ("-"), something like this
I need to scan each sequence and if a dash is found, it must be replaced with the character that is present at >80% in that particular position. In the example above, the dash in the second sequence located in the second position should be replaced by a "T", since "T" is found in 700 sequences according to the Freq value. Then, the dash in the last sequence should be replaced by an "A", since the frequency of "A" is 700 representing >80% of the total sequences.
The expected output should be something like this:
The problem is that this code will replace for the majority regardless of the 80% rule
Any help will be greatly appreciated
Moderator's Comments:
Please use CODE tags (not QUOTE, FONT, and COLOR tags), to display sample input, sample output, and code segments.
Last edited by Xterra; 06-20-2015 at 11:35 AM..
Reason: Fix tags.
What is supposed to happen if there is a dash in an input line and no other character in that position occurs more than 90% of the time? For example, if the input file had these two lines:
instead of:
what should the output file contain for these two input lines?
With well over 200 posts we would hope that you have learned something from all of the sample code we have provided. The UNIX & Linux Forums is here to help you learn how to use the tools supplied on UNIX, Linux, and other similar systems; not to act as your personal, unpaid programming staff.
Moderator's Comments:
And, please, use CODE tags for sample input, sample output, and sample code segments; not QUOTE tags, not FONT tags, not SIZE tags, and only use COLOR tags to highlight changing or unusual data or code.
That's is uncalled for. Please read my posts and then comment. I have not asked/posted anything for >2 years.
I just reread all of the last 25 posts you have submitted (going back to Oct 27th, 2011) and in those 25 posts there is not one line of code that you have shown us that you have written on your own.
It is true that your last post originally contained some code that you had written, but you edited your post and removed that code; so I can't count that. (I admit that I didn't look at the rest of your posts to try to determine if you had posted and removed other code.)
If you are unwilling to show us that you are making any attempt to write code for yourself and just ask us to write code for you; what about my statement:
Quote:
With well over 200 posts we would hope that you have learned something from all of the sample code we have provided. The UNIX & Linux Forums is here to help you learn how to use the tools supplied on UNIX, Linux, and other similar systems; not to act as your personal, unpaid programming staff.
was uncalled for? If you were in the habit of showing us your work and asking us to help fix parts that didn't work in your earlier posts, why did you stop? We are happy to help you learn why your code doesn't work. And, if you do try to write code on your own, you'll learn how to do it yourself MUCH faster.
Now, moving on to the problem at hand. With your sample input, there are 795 samples. So, to get more than 90%, you need 718 samples with a single non-dash value. For character position 47 (where there is a dash in the last input line), the counts are:
and, therefore, by your rules, the dash in that line should not be changed.
If you agree that the last line should not change, the following code seems to do what you want:
producing the output:
If you want to try this on a Solaris/SunOS system, change awk to /usr/xpg4/bin/awk.
With large input files (especially when there are very few dashes in the input), the last for loop in this script could be made much more efficient. For files with a few hundred input lines with line lengths around 50 characters, it might not be worth the time needed to improve and test it.
If you think it needs to run faster and think fixing this loop would help, feel free to look for dashes instead of evaluating each character position. If you try enhancing this script to do that and run into problems, we'll be happy to help you debug your code.
Last edited by Don Cragun; 06-20-2015 at 05:04 AM..
Reason: Fix bug in code noted in next post.
This User Gave Thanks to Don Cragun For This Post:
I apologize. Further testing showed that I have some code backwards. Testing with the sample data didn't catch it. Testing it with more data looking for corner cases made it immediately obvious that the two lines close to the end of the script:
should be changed to:
Note: This problem has now been fixed in the previous post so you if you download the entire script now, you won't have to patch it.
Last edited by Don Cragun; 06-20-2015 at 05:05 AM..
Reason: Add note.
This User Gave Thanks to Don Cragun For This Post:
hi.
I have a Fixed Length text file as input where the character positions 4-5(two character positions starting from 4th position) indicates the LOB indicator. The file structure is something like below:
10126Apple DrinkOmaha
10231Milkshake New Jersey
103 Billabong Illinois
... (6 Replies)
Hello,
For example:
12........6789101112..............20212223242526..................50 ( Positions)
LName FName DOB (Lastname starts from 1 to 6 , FName from 8 to 15 and date of birth from 21 to29)
CURTIS KENNETH ... (5 Replies)
Hi
My file has a series of rows up to 160 characters in length.
There are 7 columns for each row.
In each row, column 1 starts at position 4
column 2 starts at position 12
column 3 starts at position 43
column 4 starts at position 82
column 5 starts at... (7 Replies)
I have a list of about 200,000 lines in a text file that look like this:
1 1 120
1 80 200
1 150 270
5 50 170
5 100 220
5 300 420
The first column is an identifier, the next 2 columns are a range (always 120 value range)
I'm trying fill in the values of those ranges, and remove... (4 Replies)
I am attempting to replace positions 44-46 with YYY if positions 48-50 = XXX.
awk -F "" '{if (substr($0,48,3)=="XXX") $44="YYY"}1' OFS="" $filename > $tempfile
But this is not working, 44-46 is still spaces in my tempfile instead of YYY. Any suggestions would be greatly appreciated. (9 Replies)
I have files with hundreds of sequences with missing characters represented by a dash ("-"), something like this
I need to go sequence by sequence and if a dash is found, it should be replaced with the most common character in that particular position. Thus, in my example the dash in the second... (6 Replies)
This is what I would like to accomplish, I have an input file (file A) that consist of thousands of sequence elements with the same number of characters (length), each headed by a free text header starting with the chevron ‘>' character followed by the ID (all different IDs with different lenghts)... (9 Replies)