ZELLMECHANIK-DRESDEN/ShapeOut

Error when exporting all data to fcs

Closed this issue · 5 comments

felix-r commented

This is causing trouble on Gelas Macbook:

General information

  • Operating System: Windows 7 in VM on MacOS
  • ShapeOut version: ShapeOut 0.8.4 or 0.8.3.dev5
  • ShapeOut install: Windows 64bit installer

Steps to reproduce the problem

e.g.

  1. Open Shapeout
  2. click Export Data
  3. All event Data (.fcs)
  4. Select all Data

Expected behavior

Data is saved in .fcs file(s)

Actual behavior

There is an error message

Traceback (most recent call last):
    File "shapeout\gui\export.py", line 144, in OnBrowse
    File "shapeout\gui\export.py", line 187, in export
    File "site-packages\dclab\rtdc_dataset\export.py", line 132, in fcs
    File "site-packages\fcswrite\fcswrite.py", line 136, in write_fcs
ValueError: cannot convert float NaN to integer
paulmueller commented

Thanks for reporting this bug in such detail. The "NaN" values might be produced by e.g. computation of the Young's modulus. I already have an idea how to fix this, but I cannot reproduce this bug with my standard datasets. Could you please point me to the location of the dataset with which this error occurs (e.g. on the group drive)? Please also post any preprocessing steps you are applying.

felix-r commented

Sorry for late response. You can find the .zmso file here:
smb://biotec-files.biotec.tu-dresden.de/guck_rtdc/experiments_16_17/20171211_Gela_contrast_agents/Neutro_GdCl3_vs_Ctrl.zmso

paulmueller commented

Thank you, but I need the actual data (folders containing tdms files). I do not have access to all the data that the session refers to (e.g. \\Mac\Home\Documents\Contrast_argents\Control_samples\101017_Martin_control\Online).
I cannot reproduce the bug with the data in the folders "GdCl3" and "untreated_CTRL".

felix-r commented

Sorry again for late response. I didn't realise that some the data isn't on the server. Gela now put all the data into this folder:
\\biotec-files\guck\Members\Gela\Angela Berlin\contrast_agents_final

I hope this helps and sorry that this progresses so slowly.

paulmueller commented

I can reproduce this bug and fixed it in DC-analysis/fcswrite@a086267.
You can test it with the current development version: https://ci.appveyor.com/project/paulmueller/shapeout/build/1.0.875/job/3393b67520ft1xqj/artifacts