Section and Segment Selection

Dot notation for properties

Since sections share property names eg. a length called L it is always necessary to specify which section is being discussed.

There are three methods of specifying which section a property refers to (with each being compact in some contexts and cumbersome in others). They are given below in order of precedence (highest first).

Dot notation

This takes precedence over the other methods and is described by the syntax sectionname.varname.

Python

dendrite[2].L = dendrite[1].L + dendrite[0].L
axon.v = soma.v
print(soma.gnabar)
axon.nseg *= 3

HOC

dendrite[2].L = dendrite[1].L + dendrite[0].L
axon.v = soma.v
print soma.gnabar
axon.nseg = 3*axon.nseg

This notation is necessary when one needs to refer to more than one section within a single statement.

Stack of sections

Both HOC and Python use a section stack, but they work differently:

In HOC: The section at the top of the stack is automatically used as the default for range variables when no explicit section is specified. The syntax sectionname {stmt} temporarily pushes a section onto the stack for the duration of stmt.
In Python: The section stack exists but there is no automatic default behavior for range variables. You can query the top of the stack with n.cas() and explicitly manipulate it with push_section() and pop_section(), but this is rarely needed. Most of the time, you

can specify sec= in function calls to NEURON to specify sections to work on; this temporarily puts the specified section at the top of the stack for the function call and pops it off after.

HOC

The syntax

sectionname {stmt}

means that the currently selected section during the execution of stmt is sectionname. This method is the most useful for programming since the user has explicit control over the scope of the section and can set several range variables. Notice that after the stmt is executed the currently selected section reverts to the name (if any) it had before sectionname was seen. The programmer is allowed to nest these statements to any level.

Avoid the error:

soma L=10 diam=10

which sets soma.L, then pops the section stack and sets diam for whatever section is then on the stack.

It is important that control flow reach the end of stmt in order to automatically pop the section stack. Therefore, one cannot use the continue, break, or return statements in stmt.

Looping over sets of sections is done most often with the forall and forsec commands (see below).

In HOC, the syntax

access sectionname

defines a default section name to be the currently selected section when the first two methods (dot notation and section stack) are not in effect. There is often a conceptually privileged section which gets most of the use and it is useful to declare that as the default section. e.g.,

access soma

With this, one can, with a minimum of typing, get values of voltage, etc at the command line level. More precisely, it replaces the top of the section stack with the indicated section and so will be the permanent default section only if the section stack is empty or has only one section in it.

In general, this statement should only be used once to give default access to a privileged section. It’s bad programming practice to change the default access within anything other than an initialization procedure. The “sec { stmt }” form is almost always the right way to use the section stack.

Example:

create a, b, c, d
access a
print secname()
b {
        print secname()
        access c        // not recommended. The "go_to" of sections.
        print secname()
        d {print secname()}
        print secname()
} // because the stack has more than one section, c is popped off
print secname()     // and the second "access" was not permanent!

Python

In Python, you typically use dot notation for section properties or the sec= keyword argument for functions. The section stack can be manipulated explicitly with n.push_section() and n.pop_section(), but this is rarely necessary and not recommended except as a last resort.

The current top of the section stack can be queried with n.cas(), but unlike HOC, this doesn’t automatically apply to range variables - you must be explicit about which section you’re referencing.

There is no explicit notion of a section object in HOC but a similar effect can be obtained with the SectionRef class.

Many NEURON functions refer to a specific Section. In recent versions of NEURON, most of these either are available as section methods or take a section or segment directly. For older code or for the remaining exceptions, the active section may be specified using a sec= keyword argument.

For example:

my_iclamp = n.IClamp(0.25, sec=soma)   # better to use n.IClamp(soma(0.25)) though
num_pts_3d = n.n3d(sec=apical)         # could get the same value as an int via apical.n3d()

In Python, if no sec= keyword argument is specified, functions will use NEURON’s default Section (sometimes called the currently accessed section), which can be identified via n.cas(). The default Section is controlled by the section stack; it is initially the first Section created but entries may be pushed onto or popped off of the stack by push_section() and pop_section(). Use this only as a last resort.

However, unlike HOC, range variables in Python do not automatically use the section stack - you must always be explicit about which section you’re referencing.

pop_section()

Python

Syntax:: n.pop_section()
Description:: Take the currently accessed section off the section stack. This can only be used after a function which pushes a section on the section stack such as point_process.getloc().

Example:

from neuron import n

soma = n.Section('soma')
apical = n.Section('apical')
stims = [n.IClamp(soma(i / 4.)) for i in range(5)] + [n.IClamp(apical(0.5))]
for stim in stims:
    x = stim.get_loc()
    print(f"location of {stim} is {n.secname()}({x})")
    n.pop_section()
(Note: in this example as nseg=1, the current clamps will either be at position 0, 0.5, or 1.)

(Note: a more Pythonic way to get the location of the point-process stim is to use seg = stim.get_segment(), but this is shown as an example of using n.pop_section().)

HOC

Syntax:: pop_section()
Description:: Take the currently accessed section off the section stack. This can only be used after a function which pushes a section on the section stack such as point_process.getloc().

Example:

create soma[5]
objref stim[5]
for i=0,4 soma[i] stim[i] = new IClamp(i/4)
for i=0,4 {
        x = stim[i].get_loc()
        printf("location of %s is %s(%g)\n", stim[i], secname(), x)
        pop_section()
}

push_section()

Python

Syntax:

n.push_section(number)

n.push_section(section_name)

Description:

This function, along with pop_section() should only be used as a last resort. It will place a specified section on the top of the section stack, becoming the current section to which all operations apply. It is probably always better to use SectionRef or SectionList .

In Python, manipulating the section stack only affects what n.cas() returns and some internal functions - range variables must still be explicitly specified.

push_section(number): Push the section identified by the number returned by n.this_section(), etc. which you desire to be the currently accessed section. Any section pushed must have a corresponding n.pop_section() later or else the section stack will be corrupted. The number is not guaranteed to be the same across separate invocations of NEURON.
push_section(section_name): Push the section identified by the name obtained from sectionname(strdef). Note: at this time the implementation iterates over all sections to find the proper one; so do not use in loops.

Example:

from neuron import n

soma = n.Section('soma')
apical = n.Section('apical')

# get a number to allow pushing by number
soma_id = n.this_section(sec=soma)

# push by name
n.push_section('apical')

# push by number
n.push_section(soma_id)

# RuntimeError -- no such section
n.push_section('basal')

HOC

Syntax:

push_section(number)

push_section(section_name)

Description:

In HOC, manipulating the section stack affects the default section for range variables.

push_section(number): Push the section identified by the number returned by this_section(), etc. which you desire to be the currently accessed section. Any section pushed must have a corresponding pop_section() later or else the section stack will be corrupted. The number is not guaranteed to be the same across separate invocations of NEURON.
push_section(section_name): Push the section identified by the name obtained from sectionname(strdef). Note: at this time the implementation iterates over all sections to find the proper one; so do not use in loops.

Example:

create soma, apical

// get a number to allow pushing by number
soma { soma_id = this_section() }

// push by name
push_section("apical")

// push by number
push_section(soma_id)

Looping over sections (HOC only)

HOC provides several keywords for iterating over sections that have no direct Python equivalents. In Python, section iteration is typically done using SectionList objects or by iterating over lists of sections (such as soma.wholetree() or n.allsec()) directly.

forall

Syntax:
forall stmt

Description:
Loops over all sections, successively making each section the currently accessed section.

Within an object, forall refers to all the sections declared in the object. This is generally the right thing to do when a template creates sections but is inconvenient when a template is constructed which needs to compute using sections external to it. In this case, one can pass a collection of sections into a template function as a SectionList object argument.

The forall is relatively slow, especially when used in conjunction with issection() and ismembrane() selectors. If you are often iterating over the same sets it is much faster to keep the sets in SectionList objects and use the much faster forsec command.

The iteration sequence order is undefined but will remain the same for a given sequence of create statements.

Example:
create soma, axon, dend[3]
forall {
    print secname()
}
prints the names of all the sections which have been created.
soma
axon
dend[0]
dend[1]
dend[2]
See also

forsec, ifsec, issection(), SectionList, ismembrane()

ifsec

Syntax:
ifsec string stmt

ifsec sectionlist stmt

Description:

ifsec string stmt
Executes stmt if string is contained in the name of the currently accessed section. equivalent to if(issection(string)) stmt Note that the regular expression semantics is not the same as that used by issection. To get an exact match use ifsec ^string$

ifsec sectionlist stmt
Executes stmt if the currently accessed section is in the sectionlist.

See also

forsec, SectionList, issection()

forsec

Syntax:

forsec string stmt

forsec sectionlist stmt

Description:

forsec string stmt
equivalent to forall ifsec string stmt but faster. Note that forsec string is equivalent to forall if (issection(string)) stmt

forsec sectionlist
equivalent to forall ifsec sectionlist stmt but very fast.

These provide a very efficient iteration over the list of sections.

Example:

create soma, dend[3], axon
forsec "a" print secname()

create soma, dend[3], axon
objref sl
sl = new SectionList()
for (i = 2; i >= 0; i = i - 1) dend[i] sl.append()
forsec sl print secname()