How to print data if pattern is found in awk

Question

I have a file (A.txt ; sep=",") :

kit
Software Version =
Date And Time of Export =
Experiment Name =
Instrument Software Version =
Instrument Type = CFX
Instrument Serial Number =
Run Start Date =
Run End Date =
Run Operator =
Batch Status = VALID
Method = Novaprime
Date And Time of Export,Batch ID,Sample Name,Well,Sample Type,Status,Interpretive Result,Action*,Curve analysis
,,205920777.1,A01,Unkn-01,,,
,,neg5,A02,Neg Ctrl-01,,,
,,pos6,A03,Pos Ctrl-01,,,
,,,,,,,,,,
*reporting.

And I want to print in the "Interpretive Result" column a result if the "Sample Type" column contain a pattern : "Patient" if "Unkn" ; "NC" if "Neg" ; "PC" if "Pos". In order to have the following output (B.txt ; sep=","):

kit
Software Version =
Date And Time of Export =
Experiment Name =
Instrument Software Version =
Instrument Type = CFX
Instrument Serial Number =
Run Start Date =
Run End Date =
Run Operator =
Batch Status = VALID
Method = Novaprime
Date And Time of Export,Batch ID,Sample Name,Well,Sample Type,Status,Interpretive Result,Action*,Curve analysis
,,205920777.1,A01,Unkn-01,,Patient,,
,,neg5,A02,Neg Ctrl-01,,NC,,
,,pos6,A03,Pos Ctrl-01,,PC,,
,,,,,,,,,,
*reporting.

I've try things like :

awk -F',' -v OFS="," '(NR>1 && $5="Unkn"*){print ...}' A.txt > B.txt

But I don't manage to solved the issue. Can someone have an idea?

Thanks

$5="Unkn" is an assignment, not a comparison $5=="Unkn"

Chris Davies
– Chris Davies

2020-07-08 11:18:24 +00:00
Commented Jul 8, 2020 at 11:18 — Chris Davies
– Chris Davies, Commented Jul 8, 2020 at 11:18

Quasímodo · Accepted Answer · 2020-07-08 11:22:33Z

3

As commented above, you need == for comparison. But compare fields with regexes using the ~ operator, that is what it was made for.

pat.awk:

BEGIN{FS=OFS=","}
$5 ~ /Unkn/{$7="Patient,"}
$5 ~ /Neg/{$7="NC,"}
$5 ~ /Pos/{$7="PC,"}
{print}

$ awk -f pat.awk A.txt 
kit
Software Version =
Date And Time of Export =
Experiment Name =
Instrument Software Version =
Instrument Type = CFX
Instrument Serial Number =
Run Start Date =
Run End Date =
Run Operator =
Batch Status = VALID
Method = Novaprime
Date And Time of Export,Batch ID,Sample Name,Well,Sample Type,Status,Interpretive Result,Action*,Curve analysis
,,205920777.1,A01,Unkn-01,,Patient,,
,,neg5,A02,Neg Ctrl-01,,NC,,
,,pos6,A03,Pos Ctrl-01,,PC,,
,,,,,,,,,,
*reporting.

If you prefer the one-liner:

awk -F ',' -v OFS=',' '$5~/Unkn/{$7="Patient,"}$5~/Neg/{$7="NC,"}$5~/Pos/{$7="PC,"}1' A.txt

answered Jul 8, 2020 at 11:22

Quasímodo

19.4k4 gold badges41 silver badges78 bronze badges

Thanks ! I didn't know about the "~" assignation

nstatam
– nstatam

2020-07-08 12:00:04 +00:00
Commented Jul 8, 2020 at 12:00
@Quasimodo to make it "foolproof", you may want to consider anchoring the RegExps to the beginning of the field ($5 ~ /^Unkn/), although I acknowledge the OP only said that the column "contains" a pattern ...

AdminBee
– AdminBee

2020-07-08 13:35:39 +00:00
Commented Jul 8, 2020 at 13:35
@AdminBee Thank you for your comment. Yes, I thought better not to anchor it because of "contains".

Quasímodo
– Quasímodo

2020-07-08 17:27:13 +00:00
Commented Jul 8, 2020 at 17:27

Add a comment |

Stack Exchange Network

How to print data if pattern is found in awk

1 Answer 1

You must log in to answer this question.

Hot Network Questions

How to print data if pattern is found in awk

1 Answer 1

You must log in to answer this question.

Related

Hot Network Questions