Getting extremely large numbers like that suggests something may have gone wrong with your processing also.
Matt. From: <hcp-users-boun...@humanconnectome.org<mailto:hcp-users-boun...@humanconnectome.org>> on behalf of Timothy Coalson <tsc...@mst.edu<mailto:tsc...@mst.edu>> Date: Thursday, September 6, 2018 at 7:04 AM To: Eric Earl <e...@ohsu.edu<mailto:e...@ohsu.edu>> Cc: Darrick Sturgeon <sturg...@ohsu.edu<mailto:sturg...@ohsu.edu>>, "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>>, Anders Perrone <perro...@ohsu.edu<mailto:perro...@ohsu.edu>> Subject: Re: [HCP-Users] wb_command failed to convert text to number I have modified -cifti-convert -from-text to now accept text numbers that require double (float64) to represent them, warn about them, and replace them with the maximum of float32 (or its negation). This should cover most cases (to accept numbers that are too big even for double, I'd have to look at conversion functions other than QT's). I have made new dev_latest versions available here so you can test this change (find your OS type): http://brainvis.wustl.edu/workbench/ Note that we strongly recommend using only released versions of wb_view, as scene files saved with development versions of wb_view may contain non-final changes that may not be compatible with future releases. Tim On Wed, Sep 5, 2018 at 3:32 PM, Timothy Coalson <tsc...@mst.edu<mailto:tsc...@mst.edu>> wrote: Those numbers are too large to be represented in single precision float (float32), which stops at about 10^38 (wb_command uses float32 for virtually all data values, and the default cifti output datatype is float32). The command could be made to truncate large values at the maximum of float32 instead, possibly with a warning, or could give a different error message if conversion to double succeeds while conversion to single fails. The simple solution for now is probably to cast your output data to single precision before writing it to text (or otherwise capping it to the valid float32 exponent range). Tim On Wed, Sep 5, 2018 at 9:55 AM, Eric Earl <e...@ohsu.edu<mailto:e...@ohsu.edu>> wrote: Hi, I am trying a command like this on ten different inputs and getting a similar ERROR: …/workbench-1.3.2/bin_rh_linux64/../exe_rh_linux64/wb_command -cifti-convert -from-text …/DCANBOLDProc_v4.0.0/temp_cifti.txt …/task-SST03_Atlas.dtseries.ni<http://task-SST03_Atlas.dtseries.ni>i …/DCANBOLDProc_v4.0.0/task-SST03_DCANBOLDProc_v4.0.0_Atlas.dtseries.nii Where …/DCANBOLDProc_v4.0.0/temp_cifti.txt is the result of converting …/task-SST03_Atlas.dtseries.ni<http://task-SST03_Atlas.dtseries.ni>i to text with -cifti-convert -to-text and doing some detrending and regression then making a space-separated (within rows) text file out of those numbers. This has worked before and I am not sure why it stopped working now… Here is the actual error: ERROR: failed to convert text to number: ‘<+/- some floating-point number greater than 1 and less than 10>e+<some exponent greater than 50>’ And after that it echo’s the command it tried to run and furthermore says “Signal 127”. Any thoughts or ideas on how to resolve this would be genuinely appreciated. Warm Regards, Eric Earl, Senior Research Assistant Fair Neuroimaging Lab, Mackenzie Hall 2198 http://www.ohsu.edu/fair-lab e...@ohsu.edu<mailto:e...@ohsu.edu> Phone: 503-494-9732 Fax: 503-494-9988 Department of Behavioral Neuroscience OHSU, Mail code: L470 3181 S.W. Sam Jackson Park Rd.<https://maps.google.com/?q=3181+S.W.+Sam+Jackson+Park+Rd.+%0D%0APortland,+Oregon+97239&entry=gmail&source=g> Portland, Oregon 97239 _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org> http://lists.humanconnectome.org/mailman/listinfo/hcp-users _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org> http://lists.humanconnectome.org/mailman/listinfo/hcp-users ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
