terrytangyuan / lfda Goto Github PK
View Code? Open in Web Editor NEWLocal Fisher Discriminant Analysis in R
License: Other
Local Fisher Discriminant Analysis in R
License: Other
related to openjournals/joss-reviews#1572
I can't quickly check if the package is working, because examples don't run. See, e.g., ?lfda
. The provided examples does not run. Please check your docs / examples and provide running examples for the user.
Refer to this: http://www.ms.k.u-tokyo.ac.jp/2010/SELF.pdf
The code structure should be similar.
Please provide / copy a statement of need (from the paper?) to the readme as well (related to #26).
A minor one, there seems to be a typo in the output of lfda()
:
The trained transforming matric is:
Is it metric or matrix?
Applying the standard kernel trick allows us to obtain a non-linear extension of SELF called Kernel SELF (KSELF).
Make MetricTransformedViz() only for 3D visualization, instead of doing both lfda/klfda model building and then visualization. In other words, split the tasks.
Related to openjournals/joss-reviews#1572
The paper clearly states what the package does, but does not emphasize why it's needed. Please explain if the functionality in the lfda package is not provided by other packages, or what other needs are being addressed by your package.
White paper URL has been updated to
https://www.gastrograph.com/resources/local-fisher-discriminant-analysis-on-beer-style-clustering
After applying kmatrixGauss to the original data matrix and then apply klfda, the resulting transforming matrix cannot be applied to new data set since the dimensions don't match. I have sent an email to Dr. Sugiyama for a solution for this.
@zachmayer Do you know how long it takes for CRAN to review the package? Any suggestions before I submit?
When I use "install.packages("lfda")" to install lfda package in Rstudio with linux, there's a problem appeared. The following is problem state:
/usr/bin/ld: cannot find -lblas
/usr/bin/ld: cannot find -lgfortran
collect2: error: ld returned 1 exit status
/usr/share/R/share/make/shlib.mk:6: recipe for target 'RSpectra.so' failed
make: *** [RSpectra.so] Error 1
ERROR: compilation failed for package ‘RSpectra’
The downloaded source packages are in
‘/tmp/RtmpYbpUNY/downloaded_packages’
What should I do to solve this problem?
OS:Ubuntu 16.04 with Plasma desktop
Rstudio version 1.0.143
R version 3.2.3
Again for the JOSS review. There are six empty tests (i.e. they are not testing anything):
@zachmayer Would you be able to take a look at this? I tried several ways to install but still not passing. Similar to rgl dependency problem. How did you get managed to pass it?
The third paragraph in the paper (PCA) suddenly uses technical terms that are common in machine learning (like "feature"). From the first two paragraphs, it was not clear that or if the application of lfda is limited to machine learning. If not, I suggest a more broad / general wording, so users in disciplines like social science or psychology feel more comfortable with the content of the paper. Please either revise the wording, or introduce a link to machine learning earlier if the applicaton of lfda is only limited to ml.
For JOSS review. I have not been able to plot any of the example models in the README using the plot(model, y)
syntax. The code runs and no errors are produced, and the model
object looks ok, but plot(model, y)
does not produce anything.
What could I be missing? I have all the dependencies and suggests installed.
Some code are repetitive and can be refactored into smaller functions.
This repository does not pass two of the general checks from the JOSS review:
Related to openjournals/joss-reviews#1572
Please add a DOI to all references where available. For example, Scholkopft/Mullert or the R Journal have DOIs.
A lot of the code can be optimized, e.g. parallellization, faster eigen solver, etc, to cope with large number of classes and observations in each class.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.