Introduction to the R software for Forensim and LRmix users

25 March 2013
Introduction to the
software for
Forensim users
Hinda Haned

2 Introduction to LRmix |Avila June 2013
Part 1: general introduction to R
Part 2: introduction to the LRmix module

Part1

What is ?
is a freely available language and environment for statistical
computing and graphics
Open source means that the source code is available to all
potential users and they are free to use, modify, and re-distribute
the source code.

What is open source software?
Open source means that the source code is available to all
potential users and they are free to use, modify, and re-distribute
the source code.

What is source code?
A source code file is the original computer program written by a
programmer
It shows the logic behind a program, the strategies used for
solving various problems, and all the details needed to make the
software run

PCR simulation module in Forensim:

Source code of the module:

Open source software: Definitions
Refers exclusively to the source code and it is possible to have
support, services, documentation, and even binary versions which
are not monetarily free.
Open source = transparent
Open source  “black box”
Open source  free of charge, but most of the time it is!

Open source software: Definitions
Four essential freedoms:
1. Redistribute software without restriction
2. Access the source code
3. Modify the source code
4. Distribute the modified version of the software : for free...or
not!
Open source evolves through community cooperation:
 Community of users
 Community of developers

Open source ≠ Freeware
Freeware : A software which can be downloaded, used, and copied
without restrictions, but, no access to the source code.
There is no community and no development infrastructure around
freeware" as there is around open source software
Open source ≠ Freeware

Open source vs. Freeware: example
Open source software
Community of developers
Community of users
Users report bugs to developers
Freeware ≠ open source
Community of users
Users report bugs to Microsoft

Widely used open source software
Gimp
Mozilla Firefox
VLC Media Player
Linux
Android

Is open source secure?
“Since everyone can contribute to open source software, isn't less
reliable than commercial software?”
No ! Any change to an open source project is submitted to the
filter of a group of maintainers first
Open source software can be more secure than commercial
software:
• active communities tracking bugs and inconsistencies
• users = testers that report back to the project (mailing lists,
forums)

Open source = Free software
Example: Commercial version of the R software:
Core program is free, but the company charges for extra products
• Enterprise deployment
• Technical support
• Consulting
• Training
• …
Free as in free speech, not as in free beer!

Summary
Open source software : freedom to run, copy, distribute, study,
change and improve the software.
Open source ≠ non-commercial
“Open source promotes software reliability and quality by supporting
independent peer review and rapid evolution of source code”
- The Free Software Foundation-

Why should you use ?
Advantages
• Fast and free
• Work on the cutting edge of statistical research
• Very active user community
• Excellent for simulation, programming, computer, intensive
analyses…
• Script language: forces you to think about your analysis!
Disadvantages
• Not user-friendly
• Data preparation and cleaning might be difficult
• R-Help list : famous to be hostile !

• R is like other programming languages : C, Perl and Python
• R is particularly useful because it contains built-in mechanisms for
organizing data, running calculations, creating graphical
representations of data sets
• Researchers and engineers can improve the existing code for a
specific task, for example, the calculation of the mean, and make
these new functions available in a package
fundamentals

A package is a collection of small programs dedicated to a
specific task. You can find packages for almost anything !
fundamentals
• Statistical Genetics
• Forensic genetics
• Bayesian inference
• Computational physics
• Clinical trials
• Probability distributions
• Analysis of ecological data
• Finance
• Graphics
• Medical image analysis
• Multivariate statistics
• Statistics for the social sciences
• Analysis of spatial data
• Survival analysis
• Time series analysis
For an overview of available packages per topic:
http://cran.r-project.org/web/views/

fundamentals
 The Comprehensive R Archive Network (CRAN) : is a network
of ftp and web servers around the world that store identical,
up-to-date, versions of code and documentation for R.
 Currently, the CRAN package repository features 2343 available
packages!

The Forensim package
• Forensim is a package for the R statistical software
• Forensim is freely available
• Sources are freely available on the web
• Compiles and runs on a wide variety of UNIX platforms, Windows
and MacOS

Overview of the Forensim package
Forensim combines different features to answer some of the
identified needs in forensic genetics
Simulation tools: simulation of
data commonly encountered in
forensic casework
Statistical tools: main statistical
methods for
forensic DNA evidence
interpretation

Forensim package: simulation tools
R script
• Allele frequencies- population substructure
• DNA profiles (qualitative and quantitative data)
• DNA mixtures
User-friendly modules
• Polymorphism chain reaction: simPCR2
• Heterozygote balance (demonstration later today): Hbsimu

Forensim package: statistical tools
• Exclusion probability (RMNE)
• Likelihood ratios: General formula for likelihood ratios – LRmix
user-friendly module
• Curran et al, 1999, Balding & Buckleton, FSIG, 2009
• Random match probabilities

Forensim package: documentation
• Manual: all functions and data sets are described, examples are
given
• Detailed tutorials with practical and reproducible examples are
available online
• LRmix tutorial is distributed during the course
http://forensim.r-forge.r-project.org/

Forensim package: how to get help
 Post a message on Forensim mailing list forensim-help
 Contact me: h.haned@nfi.minvenj.nl/hi.haned@gmail.com
 Subscribe to forensimnews@gmail.com

Part 2: introduction to LRmix

(1) Install the R software

 An executable file will be downloaded
automatically.
 R.2.15.1.exe
 Simply click and follow the instructions!
 Es possible elegir Español

Press ‘next’ until...

Prepare your working folder first!
-Create a folder where you will put your cases, notes, lectures, etc
-Copy the blue R Icon in the folder

A little trick to make your life easier...

Leave it
blank
Press Apply

You are now ready to launch R
Simply click the blue Icon!

(2) Install the Forensim package
Option 1: install the package directly from the R environment
(Internet connexion)
Option 2: Install the package manually (no Internet connexion)
 Refer to LRmix tutorial online

(2) Install the Forensim package

Choose mirror
(in Spain)

Choose package forensim

Clear console for better
visibility
(nothing will be deleted)

(3) Load the Forensim library
Type the following code in the R console:
library(forensim)

(3) Load the Forensim library
library(forensim)

(4) Start LRmix
library(forensim)
LRmixTK()

Illustration of the LRmix module using the Hammer
case, Published in Gill et al, FSIG, 2007

Main LRmix interface

Input files in LRmix
Type 1: CSV files, they are comma separated files (‘,’), and the
decimal separator is the dot (‘.’)
Type 2: tab separated files, they are tab separated (‘t’, e.g. Excel),
and the dot(‘.’) is the decimal separator
Never use spaces in your column-names, or in the
sample-names (epg, or references)

Installing OpenOffice could greatly assist!!
Open office is the equivalent of Microsoft Office, except:
-It is open source
-Free of charge
-Much more efficient when it comes to visualising data
-You can visualise data much more easily than with Excel or notepad
www.openoffice.org

Main LRmix interface
(1) Load the crime-sample profile
(2) Load the references
(suspect/victim)
(3) Load your allele frequencies

(1) Load the crime-sample profiles

sampleHammer.txt

sampleHammer.txt
Import Data

Display profiles,
To make sure the data are OK
sampleHammer.txt

If everything looks good, press OK!
• You can select loci
• You can select replicates

(2) Load reference profiles
suspectHammer.txt
victimHammer.txt

You cannot see the reference
profiles

 You cannot see the reference profiles
 The program will automatically select the loci you chose in step(1)
 If there loci in the epg that are not given in the reference profile,
the program will give an error message

(3) Import the allele frequencies
identifilerUSA.txt

You may need help building frequencies file for the first
time!

 Don’t worry about the drop-out parameter now,
you will be able to carry out a sensitivity plot in the
following steps
 We will address the performance tests later on
tomorrow

(I)
Hp: Victim 1 + victim 2+ Suspect
Hd: Victim 1+ Victim 2+ Unknown
Drop-in=0.05
Drop-out=0.10
Theta=0

Dropin=0.05
Drop-out=0.10
Theta=0
Sensitivity plot

Sensitivity analysis
Drop-out ranges [Export
results]

Set-up your hypotheses
(I)
(II)
Hp: Victim 1 + Unknown+ Suspect
Hd:Victim 1+ 2 Unknowns
(III)
Hp: Victim 2 + Unknown+ Suspect
Hd: Victim 1+ 2 Unknowns

Introduction to the R software for Forensim and LRmix users

More Related Content

Similar to Introduction to the R software for Forensim and LRmix users (20)

More from hindahaned (6)

Recently uploaded (20)

Introduction to the R software for Forensim and LRmix users