Setup File:   The setup file is a series of setup-statements which may appear in any order on separate lines.  The first letter of each command is case-insensitive.

The following two setup-statements are strictly required:
Dt  delta time, in ms

Print Step Size {step size between saving data internally; resolution of resulting trace is print step size * dt}
    Note: Any line containing the word Step is acceptable here

The rest of the setup-statements are optional, but you'll probably use most of them:
Output {variable or equation names that appear in the ODE file, space or tab delimited}
This will tell all neurons that possess any of the variable names given to output those variables for all experiments.
  

Experiment experiment-statement*
experiment-statement: {experiment name} | {number of times to run last named experiment} 
Each experiment name given must have an associated Experiment file file in the same directory by the same name with the suffix .exp

Restricted Output {variable names, space or tab delimited}
This statement tells neuronetsim that you want to output specific variables or equations only on a subset of the neurons.  Each statement of this name should be followed on the next line by:
On neuron-specifier-statement
neuron-specifier-statement: Either the ID (1 indexed) of the neuron, the name of the set of neurons, or "all" or "All" to indicate all neurons or "Network" or "network" to indicate all neurons and averaged over the network as well (which will be output to "network.txt").  
 
Change variable-statement
variable-statement:
{state variable name | regular variable name} < =, *=, +=, /=, %=>  arithmetic-statement
Note: max of 1 variable statement per line; default operator is =
This changes the value of the selected variable on all neurons that have that variable for all experiments.  
The change will be reflected in the network file eventually output
arithmetic-statement:
Any normal arithmetic expression, possibly involving standard single-character operators (*,+,-,%,/,<,>), including any variables that have been defined for the neurons in the .net file (if the any neuron possesses the variable on the LHS of the equation but not all of the variables on the RHS, it will result in a crash), as well as any function-statement* (defined later)

Restricted Change neuron-specifier-statement  variable-statement
Restricts the change to the neuron or group of neurons in question.  "Network" and "network" are not valid as neuron-specifier statements here

Variable Network {A filename (not in quotes)}
The named file tells us where to find information on networks that we want to vary.  Note that, at most 1 Variable Network is allowed per file

Analyze {variable or equation name that appear in the ODE file, space or tab delimited} 
{Standard | < <Absolute | Relative> Threshold={floating-point number}> | Raw | Labeled | File < = "{filename}"> | Screen | Append | Overwrite | Rising | Falling | Narrowness={floating point number} | Detail = {Integer} | Cutoff = {floating-point number}}* 
The second line must be present; if you're happy with the defaults, just type Standard
If the threshold statement is omitted, it is assumed that there are bursts in this variable; in other words, we'll find the most burst-like aspect of the data
Any variable that is analyzed needs to have already been told to be output on the network.  
The default threshold type is absolute 
Relative thresholds should be 0-infinity.  Relative means the threshold is relative to the standard deviation of the trace (ie, 2 means that the threshold burst amplitude is twice the standard deviation of the trace)
Absolute threshold means that the amplitude of the burst must be at least that number.  If there are no apparent bursts with at least that amplitude, the trace is deemed to have no bursts

Raw means just output the statistics (tab delimited except for time ranges), expect the user to know what order they come in (perhaps by using labeled in the past)

Labeled means output the statistics with labels 
The default is Labeled

File means output the analysis to a file with the name {variablename}.analysis.txt unless another is given

Screen means output the analysis to the screen
The default is File

Append means append the analysis to the existing file with the given name

Overwrite means overwrite the file that was given, if it exists; note that these options only apply to the statisitcs; if we wanted a state trace output, 
The default is Overwrite

Standard means use all the standard options

Detail = {Integer} means output with detail level equal to integer
0 outputs: Apparent frequency/period (as determined by detected bursts), true frequency/period (as determined by fourier), # of bursts, # of interburst intervals, average burst length (in s), average burst drive, average burst amplitude, and the average value during the interbust interval.  
1 also outputs additional info on the time spent bursting, quiet, percent of time bursting, and basic statistics on bursts and interburst intervals 
2 also outputs detailed statistics on different aspects of the bursts and interburst intervals.
3 also outputs burst times and interburst times
4 also outputs a separate file = the filename+".state.txt" which contains the original trace and a trace that is 0 if in an interburst interval, 1 if in a burst
Default detail level is 1

Cutoff = {floating-point number} means cutoff that many seconds of data before sending it off to the analyzer
Default cutoff is 0

Rising indicates to expect a rising burst
 
Falling indicates to expect a falling burst (a trace whose value decreases during an actual burst).  In this case, the trace (and hence all statistics) are premultiplied by -1.  
Default is Rising

Narrowness= indicates the selectivity criteria for the burst.  The default value is 0.5; higher values will result in a smaller region being labeled a burst, while lower values will result in a much wider region being labeled a burst.  

Other parameters allowed which are barely documented and should only be used if the analysis is producing results that identify the wrong components:

minMultiplier= - Bursts must be at least this ratio of size smaller to the larger of their closest neighbors (default = 1.5; ie. this burst is at most 1.5 times smaller than the larger of the nearest bursts)

minGlobalMaxOr= - Alternatively, bursts can be this ratio of size compared to the maximum sized burst (this prevents alternating really large bursts with smaller bursts in between from negating the smaller bursts) (default = 2.0)

minGlobalMaxAnd=  - Regardless of which of the previous is fulfilled, bursts must be at least this ratio of size compared to the largest detected burst in the series (default = 6.0)

absoluteMaxAsymmetry= - Maximum asymmetry over ave allowed between left and right baselines or pits (absasymmetry = |leftAmplitude-rightAmplitude|/(leftAmplitude+rightAmplitude)) (default = 0.75)

relativeMaxAsymmetry= - Max relative asymmetry Relative symmetry = (this-ave symmmetry)/max(std symmetry,RELATIVE_MIN_STD_ASYMMETRY), default = 2.0

relativeMinStdAsymmetry= - See previous, default = 0.25

primaryLowBand= - Lowest frequency to allow in initial fourier transform (default = 0.0)

primaryHighBand= - highest frequency to allow in initial fourier transform (default = 15.0 Hz)

secondaryLowBandMultiple= - Lowest multiple of found maximum frequency to allow in the analysis trace (default = 0.3)

secondaryHighBandMultiple= - Highest multiple of found maximum frequency to allow in the analysis trace (default = 20.0)

absoluteMaxSecondaryHighBand= - Overrides the secondaryHighBandMultiple if this is lesser than that (default = 7.0)

primarySmoothingCoefficient= - the fourier trace is smoothed with a window size of primarySmoothingCoefficient/(1000*dt), default = 11, prior to finding the maximum frequency 

secondarySmoothingCoefficient= - the trace is smoothed with a window size of period/(secondarySmoothingCoefficient*dt), default = 20.0 prior to finding burst limits