#199 number format issue in tide output

[Attila Kertesz-Farkas]

Scores appear in weird format:
0x1.111e21p-1
-0x1.0bb972p-6
0x1.d033f3p-9
0x1.02d803p-8
in Tide-Windows.

input files are attached.

[William S Noble]

Note that this only happens under Windows. It works fine under MacOS.

bash-3.2$ ~/proj/crux/trunk/src/c/crux tide-index test_fasta.fasta test
INFO: Beginning tide-index.
INFO: Writing results to output directory 'crux-output'.
INFO: CPU: Genomes-MacBook-Pro-3.local
INFO: Mon Oct 6 08:18:24 PDT 2014
INFO: Running tide-index...
INFO: Writing results to output directory 'test'.
INFO: Reading test_fasta.fasta and computing unmodified peptides...
INFO: Writing decoy fasta...
INFO: Reading proteins
INFO: Precomputing theoretical spectra...
INFO: Elapsed time: 0.0117 s
INFO: Finished crux tide-index.
INFO: Return Code:0
bash-3.2$ ~/proj/crux/trunk/src/c/crux tide-search test_ms2.ms2 test
INFO: Beginning tide-search.
WARNING: The output directory 'crux-output' already exists.
Existing files will not be overwritten.
INFO: CPU: Genomes-MacBook-Pro-3.local
INFO: Mon Oct 6 08:18:38 PDT 2014
INFO: Running tide-search...
INFO: Reading index test
INFO: Reading spectra file test_ms2.ms2
INFO: Converting test_ms2.ms2 to spectrumrecords format
INFO: Sorting spectra
INFO: Running search
INFO: Elapsed time: 0.0302 s
INFO: Finished crux tide-search.
INFO: Return Code:0

[Jeff Howbert]

FWIW, that weird format is lowercase hexadecimal scientific notation; see

http://www.cplusplus.com/reference/cstdio/printf/

Jeff

On Tue, 7 Oct 2014, William S Noble wrote:

• assigned_to: Jeff Howbert --> Kaipo
• Milestone: Crux v2.0 --> post v2.0

---

** [issues:#199] number format issue in tide output**

Status: open
Milestone: post v2.0
Created: Fri Oct 03, 2014 10:28 PM UTC by Attila Kertesz-Farkas
Last Updated: Mon Oct 06, 2014 03:19 PM UTC
Owner: Kaipo

Scores appear in weird format:
0x1.111e21p-1
-0x1.0bb972p-6
0x1.d033f3p-9
0x1.02d803p-8
in Tide-Windows.

input files are attached.

---

Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/cruxtoolkit/issues/199/

To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

[Kaipo]

The issue is with line 165 on TideMatchSet.cpp:

std::ios_base::fmtflags original_flags = file->flags();
file->setf(std::ios_base::floatfield);
*file << (*i)->first.first << '\t';
file->flags(original_flags);

I committed a change that fixes it in the way I think it was meant to behave, though maybe Jeff can clarify on this code (I'm not sure we need to set flags at all?).

[Jeff Howbert]

Hi Kaipo,

I didn't we write that code, so I can't comment on the importance of any of the details. My guess is Benjamin wrote it, and perhaps Attila modified it.

Jeff

On Wed, 8 Oct 2014, Kaipo wrote:

The issue is with line 165 on TideMatchSet.cpp:

~~~~~
std::ios_base::fmtflags original_flags = file->flags();
file->setf(std::ios_base::floatfield);
file << (i)->first.first << '\t';
file->flags(original_flags);
~~~~~

I committed a change that fixes it in the way I think it was meant to behave, though maybe Jeff can clarify on this code (I'm not sure we need to set flags at all?).

---

** [issues:#199] number format issue in tide output**

Status: open
Milestone: post v2.0
Created: Fri Oct 03, 2014 10:28 PM UTC by Attila Kertesz-Farkas
Last Updated: Tue Oct 07, 2014 04:17 AM UTC
Owner: Kaipo

Scores appear in weird format:
0x1.111e21p-1
-0x1.0bb972p-6
0x1.d033f3p-9
0x1.02d803p-8
in Tide-Windows.

input files are attached.

---

Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/cruxtoolkit/issues/199/

To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

[Kaipo]

It looks like that bit of code was introduced during the merge from the p-value branch back into the trunk, maybe Attila wrote it?

[Kaipo]

Should be fixed in r16565 .