Extracting the values greater than a threshold in text file?

deepblue

2/2/23, 11:49 AM

I have a data text file as follows:

data=

arht    -0.0006 0.0001  0.0147  100.0   100.0
arht    -0.0006 0.0006  0.0133  100.0   100.0
bcmd    -0.0016 0.0010  0.0168  100.0   100.0
bcmd    -0.1012 0.0010  0.0184  100.0   100.0
bcmd     0.2011 0.0024  0.0126  100.0   100.0
.
.
.

I need to extract the lines that are greater than 0.10 and -0.10 for 2, 3, and 4 columns. For the above example, 2, 3, and 4 columns in the last two lines include values greater than 0.10 and -0.10. Then, the output file should be:

output:

bcmd    -0.1012 0.0010  0.0184  100.0   100.0
bcmd     0.2011 0.0024  0.0126  100.0   100.0

0 + 0

text-processing

elmclose

2/2/23, 12:50 PM

awk 'sqrt($2*$2)>=.1 || sqrt($3*$3)>=.1 || sqrt($4*$4)>=.1 {print}' file.txt replace file.txt by your data file.

deepblue

2/2/23, 1:11 PM

Dear @elmclose, if you can answer this question, I'll accept it. Your solution works well. Thank you very much.

Score:1

Ubuntu

terdon

2/2/23, 2:16 PM

In awk, the default action when an expression evaluates to true is to print the current line. Since you need to check the absolute value of the number, ignoring the sign, an easy trick (mentioned by user elmclose in a comment) is to check that the square root of the number times itself (which will return the number as a positive integer) is greater than the threshold. So you could simply do:

$ awk 'sqrt($2*$2)>=.1 || sqrt($3*$3)>=.1 || sqrt($4*$4)>=.1' file
bcmd    -0.1012 0.0010  0.0184  100.0   100.0
bcmd     0.2011 0.0024  0.0126  100.0   100.0

0 + 0

Score:0

Ubuntu

elmclose

2/3/23, 11:30 AM

You can also use a function within awk to check the range validity.

awk 'function fg(v) {return v<=-.1 || v>=.1 ? 1 : 0} (fg($2)+fg($3)+fg($4)){print}' mkl.txt

or more simply:

awk 'function fg(v) {return v<=-.1 || v>=.1 ? 1 : 0} fg($2)+fg($3)+fg($4)' mkl.txt

You can define any range limits by editing .1 and -.1

0 + 0