NB. Here we present menu entries according to their layout on MacOS. On other platforms the layout can slightly differ.
Python
About
g3mclass version
license
developers
Services
services preferences
Hide g3mclass
Hide others
Show all
Quit g3mclass
File
Open: open a data file
Open example : open an example data file distributed with the software
Save: save output
Open parameters: open saved modeling parameters
Save parameters: save current modeling parameters
Action
learn model
draw plots
draw heatmaps
Help
Window
Welcome
general description of g3mclass appears first-time software is opened.
Data
input data appear here once the data file is opened
Parameters
adjustable set of controls for the models, plots and heatmaps parameters (for more info, see ‘Parameters’)
Model
subtabs with the analyte names appear under main tab.
each subtab contains information on model parameters (‘par’), saved and organized in rows and columns (‘row_col’). This includes the configuration variables learned with ‘proba’ classification
‘row’, class number for each Gaussian
‘mean’, the mean value for each Gaussian
‘sd’, standard deviation for each Gaussian
‘a’, weight for each Gaussian; the sum of all Gaussians is 1.0
Bayesian information criterion (BIC)
the log likelihood function (loglike)
number of iterations (it)
whether the maximum number of iterations has been reached (false/true).
software calculated multi-cutoffs with the cutoff intervals for classes adjacent to class ‘0’, i.e., class ‘-1’ and class’1’.
‘breaks’, breaks between classes using ‘cutoff classification’
‘labels’, class description using ‘cutoff classification’
‘s. breaks’, breaks between classes using ‘stringent cutoff classification’
‘s. labels’, class description using ‘stringent cutoff classification’
software records ‘par_mod ‘, model setting parameters (for more info, see ‘Parameters’)
‘k’, number of bins in a histogram
‘khlen’, length of vector of k
‘k_var’, variable or fixed k were used (false/true)
‘thr_di’, threshold for fusing too close Gaussians
‘thr_w’, threshold for vanishing Gaussians
software records ‘creator ‘, information
‘name’, software name
‘version’, software version
‘data’, saved data directory
‘date’, file creation date/time
Model Plots
a total mixture model and GMM’s components appear here once the data file is processed
Test class
‘proba’>’cutoff’>’s.cutoff’ output and summary statistics for test classification
subtabs with the analyte names appear under the main tab.
Ref class
‘proba’>’cutoff’>’s.cutoff’ output and summary statistics for reference classification
subtabs with the analyte names appear under the main tab.
Query class
‘proba’>’cutoff’>’s.cutoff’ output and summary statistics for query classification
pulldown menu with the analyte and optional query names appears under the main tab.
Heatmaps
a graphical representation of ‘proba’>’cutoff’>’s.cutoff’ classifications
heatmap color codes: class’0’ in shade of green, classes labeled with negative (e.g., -1, -2, etc.) and positive (e.g., 1, 2, etc.) integers in shades of blue and red, respectively. The bigger the absolute integer value the more intense the corresponding color shade is.