Berkeley Data Base Utilities: Examine and interact with a EMAN2DB directory.
Usage in command line
e2bdb.py --cleanup --force --delete --all --long --short --filt='SUBSTRING' --filtexclude='SUBSTRING' --match='REGULAR_EXPRESSION' --exclude='bdb_NAME' --dump --smalldump --extractplots --check --merge --makevstack='bdb:/Home/user/project#output' --appendvstack='bdb:output' --list='abc.txt' --exlist='xyz.txt' --restore='bdb:vstack' --ppid=PPID --checkctf --step=STEP --nocache --verbose=LEVEL
e2bdb does not support MPI.
1. Clean up the database cache so files can safely be moved or accessed on another computer via NFS.
2. Delete (or at least empty) the named database(s).
e2bdb.py bdb:filename --delete
3. Only include dictionary names containing the specified substring.
e2bdb.py . --makevstack=bdb:vstack --filt=abc
A virtual stack will be generated, which contains all bdb files that have “abc” in the file name.
4. Only include dictionaries matching the provided Python regular expression.
e2bdb.py . --makevstack=bdb:vstack --match=abc
A virtual stack will be generated, which the bdb file that is named “abc”.
5. List contents of an entire database.
e2bdb.py --dump refine_01#register
6. Make a 'virtual' bdb image stack from one or more other stacks.
e2bdb.py bdb:db1 bdb:db2 --makevstack=bdb:db3
This will combine images in db1 and db2 into db3. Please note that any change in metadata of db1 or db2, there won't be changes in db3. However, if there are changes in the image data, it will appear at db3.
In order to include all stack with a common root in the name (for example, db1, db2, db3) into one virtual stack called data, use
e2bdb.py . --makevstack=bdb:data --filt=db
7. Append to or create a 'virtual' bdb image stack with its own metadata.
e2bdb.py bdb:db1 bdb:db2 --appendvstack=bdb:db3
db1 and db2 virtual stack will be appended to db3, if it exists. If not, a new db3 will be created.
8. Creates a new virtual bdb image stack by extracting images whose IDs are listed in the specified ASCII file from existed virtual stacks.
e2bdb.py bdb:#db1 bdb:#db2 --makevstack=bdb:.#db3 --list=abc.txt
Here, abc.txt is a file which contains the IDs of the selected images in each bdb file. A virtual stack db3 will be generated and contain the selected images of db1 and db2. If change makevstack to appendvstack, virtual stack of db1 and db2 will be appended to db3.
e2bdb.py bdb:#db1 bdb:#db2 --appendvstack=bdb:.#db3 --list=abc.txt
'This option allows you to select some specific images which you would like to manipulate to generate a virtual stack.'
9. Write changes in the derived virtual stack back to the original stack.
vstack is a derived image stack. –restore will allow it to write the change back to the parent image stack. This option is very useful when an operation (such as alignment) is performed on a subset of data (extracted as a virtual fact) and then the results are to be merged with the parent stack.
Output depends on the option settings.
IN EMAN2, the usage of Berkeley DB file format has been increased since it is very convenient to manipulate the metadata separately from the image data, and it is faster that other formats, such as HDF. As it has this special structure, its files are stored in a folder named EMAN2DB, which cannot manually copy or remove files in it.
Note: It is important to emphasize that one cannot manually rename or edit the files in the EMAN2DB directory. Doing so can corrupt the entire database such that programs will no longer be able to access it properly. You can safely move the directory as a whole to a different location, but otherwise it should not be modified. The e2bdb program is using to interact with the EMAN2DB directory.
Virtual stack is a stack of metadata associated with the image data. When generate a virtual stack. Unlike doing this same task with e2proc2d.py bdb:.#db1 bdb:.#db3; e2proc2d.py bdb:.#db2 bdb:.#db3, it will only generate the metadata stack. When generate a virtual stack, in the metadata, image location will be recored as data_path, original image ID will be recored as data_n, and the original stack will be recored as data_source.
By using the –list option, you can generate a virtual stack which contains metadata information of selected images. The virtual stack can be manipulate just as the original bdb files. Changes in virtual stack won't appear on the original stack. You can use the –restore option to write the change back to the original stack. It is very convenient to do so, and this functionality is expected to be very useful if one would like to use virtual stack for iteration.
Steve Ludtke, Ran Lin
Category 1:: APPLICATIONS
No Bugs known so far.