Hi guys 
,
I have a file in edf format with containing eeg records and also annotations in csv format. I only used ica when preprocessing the eeg file because real spikes other than artifact will help me when estimating. Do I have to import the Annotation file into raw file ? 
hi,
I have a file in edf format with containing eeg records and also annotations in csv format. I only used ica when preprocessing the eeg file because real spikes other than artifact will help me when estimating. Do I have to import the Annotation file into raw file ?
yes with:
annotations = mne.read_annotations(‘file.csv’)
raw.set_annotations(annotations)
thanx a lot @agramfort but now ı m getting error
KeyError: ‘onset’ 
omg annot, maps the time and what happen in this time. ı have to combine with raw file
I am enlightened 
“they provide a mapping between times during an EEG/MEG recording and a description of what happened at those times. In other words, they associate a when with a what”
-Mne tutorials
raw.annotations.onset → gives
array([], dtype=float64)
so your annotations are not aligned with the time of the EEG data?
A
nope I am using A dataset of neonatal EEG recordings with seizures annotations | Zenodo this open dataset for my project there are 3 annotations file in csv format and also patients eeg records 
have you understood how the annotations are stored? basically you need to know the
onsets, durations of each annotation and potential a text description
Alex
Thanx Mr. Alexandre for your comment. I am researching at the topics you mentioned now 
