Fast-aggregate¶
The fast-aggregate program aggregates trained ACP and VAP models, it does not work for precomputed data or TCP. It will merge multiple models into a large aggregated model, but it does no sanitation at all. Use it for speeding up training of larger ACP and VAP models, especially when using LibSvm which has a much longer training time for a given training set size.
Table of Contents
Parameters¶
The full usage menu can be retrieved by running command:
> java -jar cpsign-[version].jar fast-aggregate
fast-aggregate
SYNOPSIS
------------------------------------------------------------------------------------------
fast-aggregate [options]
fast-aggregate @/tmp/runconfigs/parameters.txt [options]
fast-aggregate @C:\Users\User\runconfigs\parameters.txt [options]
DESCRIPTION
------------------------------------------------------------------------------------------
The fast-aggregate program joins partially trained predictors into the final aggregated
one. This program facilitate a high level way of distributing the training step for ACP
and VAP predictors. Use the --splits flag in the 'train' program to train a partial
predictor, and then use fast-aggregate to merge everything together.
OPTIONS
------------------------------------------------------------------------------------------
Input:
* -m | --modelfiles [[URI | path] [URI | path] ..]
A list (space or comma-separated) of models that should be aggregated. It is
allowed to give directories, glob patterns (with wildcard characters), explicit
files or URIs. Note that models can be a mix of non-encrypted and encrypted models.
-af | --accept-fail
Accept failure if a model cannot be added to the aggregated model (i.e. if model is
of wrong type etc.). Default is to fail execution
Output:
* -mo | --model-out [path]
Model file to generate
General:
* --license [URI | path]
Path or URI to license file
-h | --help | man
Get help text
--short
Use shorter help text (used together with the --help argument)
--logfile [path]
Path to a user-set logfile, will be specific for this run
--silent
Silent mode (only print output to logfile)
--echo
Echo the input arguments given to CPSign
--seed [integer]
Set this flag if an explicit RNG seed should be used in tasks that require a RNG
(randomization of training data, splitting in cross-validation, learning algorithms
etc). Not used by all programs.
--progress-bar
Add a Progress bar in the system error output
--progress-bar-ascii
Add a Progress bar in ASCII in the system error output
--time
Print wall-time for all individual steps in execution
------------------------------------------------------------------------------------------
Example usage¶
> java -jar cpsign-[version].jar fast-aggregate \
--license /path/to/Standard-license.license \
-m \
models/acp_model_reg_1.cpsign \
models/acp_model_reg_2.cpsign \
--model-out \
/tmp/aggregated_reg.cpsign \
Running with Standard License registered to [Name] at [Company]. Expiry
date is [Date]
Starting to aggregate models..
Successfully aggregated 2 models.
Aggregated model saved at:
/private/tmp/aggregated_reg.cpsign