All shell commands

help

help [command]

Display help.

Positional arguments

Name	Type	Description
`command`	`{load,unload,rename,...}`	Command for which to display help

load

load <filename +> [--name +] [--transform +] [--shader 1]

Load a file, which can be local or remote.

Positional arguments

Name	Type	Description
`filename`	`list[uri]`	Filename(s) with protocols

Options

Flag	Type	Description	Default
`-n`, `--name`	`list[str]`	A name for the image layer(s).	`basename(filename)`
`-t`, `--transform`	`uri` \| `list[float]`	Apply a transform . Can be a path to a transform file or a flattened matrix.
`-s`, `--shader`	`uri` \| `str`	The name of a builtin shader, or the path to a lookup table.

Paths and URLs

Each path or url may be prepended by:

A layer protocol that indicates the kind of object that the file contains.
- Examples: volume://, labels://, tracts://.
A format protocol that indicates the exact file format.
- Examples: nifti://, zarr://, mgh://.
An access protocol that indicates the protocol used to access the files.
- Examples: https://, s3://, dandi://.

All of these protocols are optional. If absent, a guess is made using the file extension.

Examples
Absolute path to local file	`/absolute/path/to/mri.nii.gz`
Relative path to local file	`relative/path/to/mri.nii.gz`
Local file with format hint	`mgh://relative/path/to/linkwithoutextension`
Remote file	`https://url.to/mri.nii.gz`
Remote file with format hint	`zarr://https://url.to/filewithoutextension`
File on dandiarchive	`dandi://dandi/<dandiset>/sub-<id>/path/to/file.ome.zarr`

Layer names

Neuroglancer layers are named. The name of the layer can be specified with the --name option. Otherwise, the base name of the file is used (that is, without the folder hierarchy).

If multiple files are loaded and the --name option is used, then there should be as many names as files.

Transform

A spatial transform (common to all files) can be applied to the loaded volume. The transform is specified with the --transform option, which can be a flattened affine matrix (row major) or the path to a transform file.

Type help transform for more information.

Shader

A shader (= colormap, common to all files) can be applied to the loaded volume. The shader is specified with the --shader option, which can be the name of a colormap, the path to a LUT file, or a snippet of GLSL code.

Type help shader for more information.

unload

unload <layer+>

Unload layers

Positional arguments

Name	Type	Description
`layer`	`list[str]`	Name(s) of layer(s) to unload.

rename

unload <src> <dst>

Rename a layer.

Positional arguments

Name	Type	Description
`src`	`str`	Name of layer to rename.
`dst`	`str`	New name for the layer.

world_axes

Implicit Explicit

world_axes <dst *> [--source *] [--print]

world_axes [--destination *] [--source *] [--print]

Neuroglancer is quite flexible in the sense that it does not have a predefined "hard" coordinate frame in which the data lives. Instead, arbitrary "model" axes can be defines, in terms of an affine transformation of "native" axes.

By default, most formats use (x, y, z) as their model axes, although they may not have predefined anatomical meaning. NIfTI files do have an anatomical convention, under which axes x, y, z map to the right, anterior, superior sides of the brain.

This function allows native axes to replaces by more anatomically meaningful names, by leveraging neuroglancer's transforms. In order to allow these transforms to be undone, and new data to be loaded without introducing conflicting coordinate frames, we store the mapping inside local annotation layers named __world_axes_native__ and __world_axes_current__.

Positional arguments

Name	Type	Description
`dst`	`str`	New axis names.

Options

Flag	Type	Description	Default
`-d`, `--dst`, `--destination`	`list[str]`	New axis names.
`-s`, `--src`, `--source`	`list[str]`	Native axis names.	`[x, y, z]`
`-p`, `--print`		Print result.

rename_axes

Implicit Explicit

rename_axes <dst *> [--source *] [--print]

rename_axes [--destination *] [--source *] [--print]

Rename axes.

Positional arguments

Name	Type	Description
`dst`	`str`	New axis names.

Options

Flag	Type	Description
`-d`, `--dst`, `--destination`	`list[str]`	New axis names.
`-s`, `--src`, `--source`	`list[str]`	Old axis names.
`-p`, `--print`		Print result.

space

space [{radio,neuro,default}] [--layer 1]

This function rotates and orients the cross-section plane such that

Visual axes are pointed according to some defined convention (radio or neuro)
The cross section is aligned with either the model coordinate frame, or one of the layer's voxel space.

Note

When this function is used, the displayed axes are always reset to x, y, z (or their world names if world_axes has been used).

Examples
`space radio`	Orients the cross section such that the x axis points to the left of the first quadrant (radiological convention).
`space neuro`	Orients the cross section such that the x axis points to the right of the first quadrant (neurological convention).
`space --layer <LAYER>`	Aligns the cross-section with the voxels of the designated layer, while keeping the existing radio or neuro convention.
`space --layer world`	Aligns the cross-section with the canonical model space, while keeping the existing radio or neuro convention.
`space radio --layer <LAYER>`	Aligns the cross-section with the voxels of the designated layer, and uses the radiological convention

Positional arguments

Name	Type	Description
`space`	`{radio,neuro,default}`	Space.

Options

Flag	Type	Description	Default
`-l`, `--layer`	`str`	If the name of a layer, align the cross-section with its voxel grid. If "world", align the cross section with the canonical axes.

transform

transform <transform *> [--layer +] [--inv] [--reset] [--mov] [--fix]

Apply a transform.

Positional arguments

Name	Type	Description
`transform`	`uri` \| `list[float]`	Path to transform file or flattened transformation matrix (row major).

Options

Flag	Type	Description	Default
`-l`, `--layer`	`list[str]`	Name(s) of layer(s) to transform	All layers
`-i`, `--inv`		Invert the transform before applying it
`-r`, `--reset`		Reset the layer's transform before applying the new one
`-m`, `--mov`	`uri`	Moving image (required by some formats)
`-f`, `--fix`	`uri`	Fixed image (required by some formats)

save_transform

save_transform <output *> [--layer +] [--format 1]

Save the current transform

Positional arguments

Name	Type	Description
`output`	`list[uri]`	Path to output transform(s).

Options

Flag	Type	Description	Default
`-l`, `--layer`	`list[str]`	Name(s) of layer(s) to save.
`-f`, `--format`	`{afni,fsl,itk,lta}`	Format to save into.

channel_mode

channel_mode <{local,channel,global}> [--layer +] [--dimension +]

	In neuroglancer, axes can be interpreted in three different ways
global	The axis is common to all layers and can be navigated (default for axes that have spatial or temporal units)
local	The axis is specific to a layers and navigation is local (dimension names end with `'`)
channel	The axis is specific to a layer; enables multi-channel shading (dimension names end with `^`)

Note

When an axis is local ('), only one channel can be shown at once.
When an axis is a channel (^), all channels can be mixed into a single view using a multi-channel shader.
When an axis is global, navigation is linked across layers.

Positional arguments

Name	Type	Description
`mode`	`{local,channel,global}`	How to interpret the channel (or another) axis.

Options

Flag	Type	Description	Default
`-l`, `--layer`	`list[str]`	Name(s) of layer(s) to which the command is applied.
`-d`, `--dimension`	`str`	Name(s) of axes to transform.

shader

shader <shader> [--layer +] [--layer-type +]

Applies a colormap, or a more advanced shading function to all or some of the layers.

The input can also be the path to a (local or remote) freesurfer LUT file.

Positional arguments

Name	Type	Description
`shader`	`str`	Shader name or GLSL shader code.

Options

Flag	Type	Description	Default
`-l`, `--layer`	`list[str]`	Name(s) of layer(s) to which the shader is applied.
`-t`, `--type`	`{image,segmentation,annotation,...}`	Layer type(s) to which the shader is applied.

display

display <dimension *>

Neuroglancer is quite flexible in the sense that it does not have a predefined "hard" coordinate frame in which the data is shown. Instead, the red, green, blue "visual" axes can be arbitrarily mapped to any of the existing "model" axes.

By default, most formats use (x, y, z) as their model axes, although they may not have predefined anatomical meaning. NIfTI files do have an anatomical convention, under which axes x, y, z map to the right, anterior, superior sides of the brain.

In order to show data in a frame that is standard in the neuroimaging field, we therefore map the visual axes red, green, blue to the model axes x, y, z when loading data in an empty scene. This mapping is also enforced every time the command space is used.

The display command can be used to assign a different set of model axes to the visual axes.

Positional arguments

Name	Type	Description
`dimension`	`str`	Dimensions to display.

layout

layout [{xy,yz,xz,3d,xy-3d,yz-3d,xz-3d,4panel}]                       \
       [--stack {row,column}] [--layer +] [--flex] [--row] [--column] \
       [--append *] [--assign *] [--insert *] [--remove *]

Change the viewer's layout (i.e., the quadrants and their layers)

	Neuroglancer has 8 different window types
`xy`	XY cross-section
`yz`	YZ cross-section
`xz`	XZ cross-section
`xy-3d`	XY cross-section in a 3D window
`yz-3d`	YZ cross-section in a 3D window
`xz-3d`	XZ cross-section in a 3D window
`4panel`	Four quadrants (XY, XZ, 3D, YZ)
`3d`	3D window

It is possible to build a user-defined layout by stacking these basic windows into a row or a column — or even nested rows and columns — using the --stack option. The --layer option allows assigning specific layers to a specific window. We also define --append and --insert to add a new window into an existing stack of windows.

Positional arguments

Name	Type	Description
`layout`	`{xy,yz,xz,3d,xy-3d,yz-3d,xz-3d,4panel}`	Layout.

Options

Flag	Type	Description	Default
`-s`, `--stack`	`{row,column}`	Stack direction
`-l`, `--layer`	`list[str]`	Layer(s) to include
`-f`, `--flex`	`float`	Flex	`1`
`-a`, `--append`	`optional[list[int]]`	Append to existing (nested) layout
`-x`, `--assign`	`optional[list[int]]`	Assign into existing (nested) layout
`-i`, `--insert`	`optional[list[int]]`	Insert in existing (nested) layout
`-r`, `--remove`	`optional[list[int]]`	Remove from an existing (nested) layout
`--row`		Alias for `--stack row`
`--col`, `--column`		Alias for `--stack column`

state

state [--no-print] [--save 1] [--load 1] [--url] [--open] [--instance {ng,linc}]

Return the viewer's state.

Options

Flag	Type	Description
`--no-print`		Do not print the state.
`-s`, `--save`	`uri`	Save JSON state to this file.
`-l`, `--load`	`uri` \| `str`	Load JSON state from this file. Can also be a JSON string or a URL..
`-u`, `--url`		Load (or print) the url form of the state.
`-o`, `--open`		Open the url (if `--url`) or viewer (otherwise)
`-i`, `--instance`	`{ng,linc}`	Link to this neuroglancer instance

move

move <coord *> [--dimensions +] [--unit 1] [--absolute] [--reset]

Move cursor

Positional arguments

Name	Type	Description
`coord`	`list[float]`	Cursor coordinates. If None, print current one.

Options

Flag	Type	Description
`-d`, `--dimensions`	`list[str]`	Axis name for each coordinate (can be compact)
`-u`, `--unit`	`str`	Coordinates are expressed in this unit
`-a`, `--abs`, `--absolute`		Move to absolute position, rather than relative to current
`-r`, `--reset`		Reset coordinates to zero

zoom

zoom [factor] [--reset]

Zoom by a factor [default: x2]

Positional arguments

Name	Type	Description	Default
`factor`	`float`	Zoom factor.	`2`

Options

Flag	Type	Description	Default
`-r`, `--reset`		Reset zoom level to default.

unzoom

unzoom [factor] [--reset]

Unzoom by a factor [default: ÷2]

Positional arguments

Name	Type	Description	Default
`factor`	`float`	Inverse zoom factor.	`2`

Options

Flag	Type	Description	Default
`-r`, `--reset`		Reset zoom level to default.

zorder

zorder <layer *> [--up] [--down]

Modifies the z-order of the layers.

In neuroglancer, layers are listed in the order they are loaded, with the latest layer appearing on top of the other ones in the scene.

Counter-intuitively, the latest/topmost layer is listed at the bottom of the layer list, while the earliest/bottommost layer is listed at the top of the layer list. In this command, layers should be listed in their expected z-order, top to bottom.

There are two ways of using this command:

Provide the new order (top-to-bottom) of the layers
Provide a positive (--up) or negative (--down) number of steps by which to move the listed layers. In this case, more than one up or down step can be provided, using repeats of the option.

Example

-vvvv moves downward 4 times
-^^^^ moves upwards 4 times

Positional arguments

Name	Type	Description
`layer`	`list[str]`	Layer name(s).

Options

Flag	Type	Description	Default
`-^`, `-u`, `--up`		Move upwards.
`-v`, `-d`, `--down`		Move downwards.

cd

cd <path>

Change directory.

ls

ls <path> [-l,--long] [-h,--hidden]

List files.

pwd

pwd

Path to working directory.

stdin

stdin <file>

Set input stream.

stdout

stdout <file>

Set output stream.

stderr

stderr <file>

Set error stream.

exit (quit)

exit

quit

Exit nglocal.

Examples
`world_axes right anterior superior`	`x > right, y > anterior, z > superior`
`world_axes ras`	`x > right, y > anterior, z > superior`
`world_axes u v w --src z y x`	`x > w, y > v, z > u`