# \[mne-python\] strange behavior of raw.save() **URL:** \ **Category:** Mailing List Archive (read-only)\ **Tags:** list-archive\ **Created:** [September 8, 2012, 8:04pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484 "2012-09-08T20:04:53Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 8:04pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/1 "2012-09-08T20:04:53Z")
Hi all, on saving a raw file using the instance method save() I get the following error message. [http://pastebin.com/g66H4svV](http://pastebin.com/g66H4svV) After that the file is not readable anymore ( the same error message will appear as Raw.\_\_init\_\_() is called. ) If I use an alternative file name this does not happen. a) what is the reason for that / what does it actually tell, and b) shouldn't that rather throw an exception than messing up the original file? Cheers, Denis ---
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 9:39pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/2 "2012-09-08T21:39:08Z")
Hi Denis, Are you by any chance saving the raw file to the same file from where you read it? I.e., something like raw = Raw('test\_raw.fif') .. raw.save('test\_raw.fif') Unless you use preload=True, this won't work since raw.save() will read from the file and at the same time overwrite it. We should add a check such that it throws and exception with a explanatory error message if one tries to do this. Martin ---
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 9:47pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/3 "2012-09-08T21:47:27Z")
Thanks Martin, that must be it. I actually I was testing a new command line tool of mine for mass riff export / augmentation / conversion And in one of the use cases I simply wanted to change the info structure without loading the data. So this means I can either preload the data or work around it by e.g. iterative renaming. To prevent users from tnings like that one could add something like raw.save(fname) (...) if fname == self.raw.info['filename'] and not self.preload:     raise ValueError("Invalid operation "                             "Either load data or use alternative file name") Wdyt? Cheers, Denis 2012/9/8 Martin Luessi \: ---
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 10:01pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/4 "2012-09-08T22:01:27Z")
Ok, I tried the following: I added if fname == self.info['filename'] and not hasattr(self,'\_data'):             raise ValueError('Invalid operation. Either read data with preload '                                     'option or use alternative file name') to Raw.save Hower, in case the data is preloaded the file still breaks... 2012/9/8 Denis-Alexander Engemann \: ---
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 10:03pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/5 "2012-09-08T22:03:27Z")
Ok. self.\_preload 2012/9/9 Denis-Alexander Engemann \: ---
**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\ **Post date:** [September 8, 2012, 10:23pm UTC](https://mne.discourse.group/t/mne-python-strange-behavior-of-raw-save/484/6 "2012-09-08T22:23:14Z")
Ok, self.\_data == True or self.\_preodaded == True does not make a difference. Is it at all possible to write make a ras.save using the identical fid? Otherwise we should just add: if fname == self.info['filename']:             raise RuntimeError('Invalid operation. Use an alternative file name') 2012/9/9 Denis-Alexander Engemann \