DataFolder#

class nextnanopy.DataFolder(fullpath)#

Bases: object

Helper for the navigation in the output folder of the nextnano simulations.

Lists the files the directory holds and a DataFolder for each subfolder, built recursively when the object is made – so it is a snapshot of the tree as it stood then.

A subfolder can be reached three ways:

folder.folders['results']
folder.go_to('results', 'bias_0')
folder.results.bias_0

The last needs a name that is a usable Python identifier and that DataFolder is not already using: a clashing name such as files warns and gets no attribute, and a name holding a space or a dot gets one that no dotted expression can reach. go_to works whatever the name.

Parameters:

fullpathstr or pathlib.Path: Path to the directory.

Attributes:

fullpathstr or pathlib.Path: Path to the directory.
foldersDictList: A DataFolder for each direct subfolder, keyed by subfolder name.
fileslist of str: Path of each file lying directly in the directory, in name order. Files held by the subfolders are not among them.
namestr: Name of the directory itself, without the path leading to it. Read-only.

find(template, deep=False)#

Return the files whose name contains template.

Parameters:

templatestr: Substring to look for. Matched against the file name alone, not the folders leading to it, and taken literally – case matters and there is no globbing, so *.dat finds nothing while .dat finds every one.
deepbool, default=False: Whether to search the subfolders as well, at every depth.

Returns:

list of str: Path of every file that matched, this directory’s first and then each subfolder’s. Empty when nothing matched.

Examples

>>> folder = DataFolder('output')
>>> folder.find('bias')
['output/a_bias.dat', 'output/b_bias.dat']

Reaching into the subfolders as well:

>>> folder.find('bias', deep=True)
['output/a_bias.dat', 'output/b_bias.dat', 'output/sweep/z_bias.dat']

The template is a plain substring, so a glob matches nothing and case matters:

>>> folder.find('*.dat')
[]
>>> folder.find('BIAS')
[]

find_multiple(templates, deep=False)#

Return the files whose name contains every one of templates.

Parameters:

templateslist of str: Substrings that must all appear in the file name, in any order. Each is matched the way find matches its own: against the name alone, taken literally, case included. An empty list matches every file.
deepbool, default=False: Whether to search the subfolders as well, at every depth.

Returns:

list of str: Path of every file that matched, this directory’s first and then each subfolder’s. Empty when nothing matched.

See also

find: Search for a single substring.

Notes

Pass a list, never a bare string. A string is iterated character by character, so find_multiple('bias') asks for the letters b, i, a and s in any order rather than for the word.

Examples

>>> folder = DataFolder('output')
>>> folder.find_multiple(['bias', '.dat'], deep=True)
['output/a_bias.dat', 'output/b_bias.dat', 'output/sweep/z_bias.dat']

Every template has to appear, so adding one narrows the result:

>>> folder.find_multiple(['bias', 'a_'])
['output/a_bias.dat']

go_to(*args)#

Follow a path down from this directory and return what is there.

Parameters:

*args: Names to descend through, one per level, joined onto fullpath. With none, the directory itself is returned.

Returns:

DataFolder or str: A new DataFolder when the path names a directory, the path itself when it names a file.

Raises:

ValueError: If the path names neither a file nor a directory.

Notes

The directory is scanned again, so the object handed back is a new one and not the DataFolder already sitting in folders. It therefore sees files written since, at the cost of walking that part of the tree once more.

Examples

>>> folder = DataFolder('output')
>>> sweep = folder.go_to('sweep')
>>> sweep.name
'sweep'

A file comes back as its path, ready to hand to DataFile:

>>> folder.go_to('sweep', 'z_bias.dat')
'output/sweep/z_bias.dat'

show_tree(with_files=True, deep=True)#

Print the folder tree.

Parameters:

with_filesbool, default=True: Whether to list the files as well as the folders.
deepbool, default=True: Whether to descend past the direct subfolders.

read_sweep_infodict()#

Read the sweep information stored in this directory.

Returns:

dict of {strdict}: The swept variable combination of each simulation, keyed by its output folder, as Sweep.execute wrote it to sweep_infodict.json.

Raises:

FileNotFoundError: If the directory holds no sweep_infodict.json, so it is not the output of a sweep.