Skip to content
R package to perform a matching analysis of two shoeprints
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
R
data
inst/bin
man
.Rbuildignore
.gitignore
BinaryFiles
DESCRIPTION
NAMESPACE
README.md
shoeprintr.Rproj

README.md

Shoeprint Boosted

How-To

  1. Make sure you have the devtools package installed
  2. From R, run devtools::install_github("CSAFE-ISU/shoeprintr")
  3. (If running on the CSAFE server or another machine that restricts executing binary files) From a Terminal, run chmod +x /home/username/R/x86_64-pc-linux-gnu-library/3.4/shoeprintr/bin/lin64/pmc replacing the path as necessary
  4. From R, Load shoeprintr by typing library(shoeprintr)
  5. Run the following sample code, modifying the input files and parameters as necessary:
    ## Load in the input and the reference images
    data(input_example)
    data(reference_example)
    
    ## Transform all the points to have (0,0) at lower left corner.
    print_in <- leftcorner_cent(input_example)
    print_ref <- leftcorner_cent(reference_example)

    ## Perform Print Match
    print_stats <- match_print(print_in, print_ref,
			       ncross_in_bins = 30, xbins_in = 20, ncross_in_bin_size = 1,
			       ncross_ref_bins = NULL, xbins_ref = 30, ncross_ref_bin_size = NULL,
			       eps = .75, seed = 1, num_cores = parallel::detectCores(), 
			       plot = TRUE, verbose = TRUE)
    print_stats
  1. (Optional) We provide the pmc binary with the R package. If you would like to update this binary to use the latest development version, follow the instructions at the bottom of this README file.

Deliverables

  1. The cleaned up, commented, documented code as an R package. The R package will enable easy sourcing of the utility functions and seamless documentation
  2. A 10-15 slide rmarkdown presentation walking through 1) the changes, 2) the speed improvement, 3) live demo
  3. A README for the package which walks through its installation and use
  4. A spreadsheet of profiling results showing the speedup with various parameters compared to the original script

PMC Instructions

  1. Clone the pmc library
  2. Go into pmc directory
    • cd pmc
  3. Compile the library
    • make
  4. Replace the binary corresponding to your platform in inst/bin/ with the newly compiled pmc binary
You can’t perform that action at this time.