Home  
  Spike Train Toolkit
     - Primer
     - Introduction
     - Toolkit Wiki
     - Download
     - Demonstrations
     - Input Format
     - Input Structures
     - Output Structures
     - Information Options
     - Entropy Options
     - Function Reference
     - Version History
     - License
     - Contribute
  Analysis Overview
     - Analyze Data
     - Your Submissions
  Neurodatabase
  BrainML Models
  LNI Homepage
 
User ID
Password


  Spike Train Analysis Toolkit Options for Information Methods



Options and parameters are passed to the toolkit functions by way of a specialized data structure. The members of this data structure are described below. Items that only apply to multineuron analysis are in pink.

Options

Default selection in blue.

Name Description Options Method
direct metric binless ctwmcmc
entropy_estimation_method Array of entropy estimation methods plugin Plug-in × × × ×
tpmc Treves-Panzeri-Miller-Carlton
jack Jackknife
ma Debiased Ma bound
bub Best upper bound
chaoshen Chao-Shen
ww Wolpert-Wolf
nsb Nemenman-Shafee-Bialek
variance_estimation_method Array of variance estimation methods nsb_var Variance in the NSB estimate × × × ×
jack Jackknife
boot Bootstrap
unoccupied_bins_strategy
See matrix2hist2d for description of options.
Strategy for dealing with unoccupied bins -1 Ignore unoccupied bins   × ×  
0 Use an unoccupied bin only if its row and column are occupied
1 Use all bins
sum_spike_trains
See directbin for description of options.
Should simultaneous spike trains be summed? 0 Do not sum across trials ×      
1 Sum across trials
permute_spike_trains
See directbin for description of options.
Should permuted versions of simultaneous spike trains be considered identical? 0 Take into account spike train origin ×      
1 Disregard spike train origin
metric_family
See metricdist for description of options.
Which family of metrics to use 0 Dspike   ×    
1 Dinterval
parallel
See metricdist for description of options.
Whether or not to use the "all-parameter" method 0 Single parameter (default if shift_cost only has one element)   ×    
1 All parameter (default if shift_cost has multiple elements)
warping_strategy
See binlesswarp for description of options.
Warping strategy 0 Linear scaling     ×  
1 Uniform spacing
stratification_strategy
See binlessinfo for description of options.
Stratification strategy 0 Single stratum     ×  
1 Stratum for each spike count
2 Stratum for each spike count; all spike trains with more than max_embed_dim-min_embed_dim spikes go into a single stratum
singleton_strategy
See binlessinfo for description of options.
Singleton counting strategy 0 Ignore     ×  
1 Include
legacy_binning
See directbin for description of options.
Legacy binning flag 0 Use the current binning method. ×      
1 Use the legacy binning method.
h_zero
See ctwmcmctree for description of options.
Deterministic node flag 0 Weight deterministic nodes.       ×
1 Do not weight deterministic nodes.
tree_format
See ctwmcmctree for description of options.
CTW tree format none Do not output tree(s) (see ctwmcmcbridge for details).       ×
cell Output tree(s) as cell array.
struct Output tree(s) as struct array.
recording_tag
See binlessembed for description of options. You do not need to set this option if you use the top level function binless.
Recording Tag episodic Data is episodic (e.g., spike trains).     ×  
continuous Data is continuous (e.g. LFP).

Parameters

Name Description Type Range Default Method
direct metric binless ctwmcmc
start_time Starting time for analysis in seconds double < end_time max of all spike train start times × × × ×
end_time Ending time for analysis in seconds double > start_time min of all spike train end times × × × ×
counting_bin_size Counting bin size in seconds double > 0 end_time-start_time ×     ×
words_per_train Number of words per spike train in each trial integer > 0 1 ×      
letter_cap Cap the maximum letter allowed in a bin integer > 0 ×     ×
shift_cost Cost metric in 1/seconds (may be a vector of such values) double ≥ 0 1/(end_time-start_time)   ×    
label_cost Cost of changing a spike label (may be a vector of such values) double ≥ 0 and ≤ 2 0   ×    
clustering_exponent Clustering exponent double -2   ×    
start_warp Starting time for warped spike trains double < end_warp -1     ×  
end_warp Ending time for warped spike trains double > start_warp 1     ×  
min_embed_dim Minimal embedding dimension for episodic data integer max_embed_dim 1     ×  
max_embed_dim Maximal embedding dimension for episodic data integer max_embed_dim 2     ×  
cont_min_embed_dim Minimal embedding dimension for continuous data integer cont_max_embed_dim 0     ×  
cont_max_embed_dim Maximal embedding dimension for continuous data integer cont_max_embed_dim 2     ×  
beta KT ballast parameter double > 0 1/(largest value in binned data plus one)       ×
gamma Weighting between tree node and its children double > 0 and < 1 0.5       ×
max_tree_depth Maximum tree depth integer > 0 1000       ×
memory_expansion Ratio by which to reallocate tree memory double ≥ 1 1.61       ×
nmc Number of Monte Carlo samples integer ≥ 0 199       ×
mcmc_iterations Absolute number of MCMC iterations integer > 0 100       ×
mcmc_max_iterations Maximum number of MCMC iterations integer mcmc_iterations 105       ×
mcmc_min_acceptances Minimum number of MCMC acceptances integer > 0 20       ×



 Webmaster

Weill Medical College of Cornell University