+
+(define-public python-autograd
+ (let* ((commit "442205dfefe407beffb33550846434baa90c4de7")
+ (revision "0")
+ (version (git-version "0.0.0" revision commit)))
+ (package
+ (name "python-autograd")
+ (home-page "https://github.com/HIPS/autograd")
+ (source (origin
+ (method git-fetch)
+ (uri (git-reference
+ (url home-page)
+ (commit commit)))
+ (sha256
+ (base32
+ "189sv2xb0mwnjawa9z7mrgdglc1miaq93pnck26r28fi1jdwg0z4"))
+ (file-name (git-file-name name version))))
+ (version version)
+ (build-system python-build-system)
+ (native-inputs
+ `(("python-nose" ,python-nose)
+ ("python-pytest" ,python-pytest)))
+ (propagated-inputs
+ `(("python-future" ,python-future)
+ ("python-numpy" ,python-numpy)))
+ (arguments
+ `(#:phases (modify-phases %standard-phases
+ (replace 'check
+ (lambda _
+ (invoke "py.test" "-v"))))))
+ (synopsis "Efficiently computes derivatives of NumPy code")
+ (description "Autograd can automatically differentiate native Python and
+NumPy code. It can handle a large subset of Python's features, including loops,
+ifs, recursion and closures, and it can even take derivatives of derivatives
+of derivatives. It supports reverse-mode differentiation
+(a.k.a. backpropagation), which means it can efficiently take gradients of
+scalar-valued functions with respect to array-valued arguments, as well as
+forward-mode differentiation, and the two can be composed arbitrarily. The
+main intended application of Autograd is gradient-based optimization.")
+ (license license:expat))))
+
+(define-public python2-autograd
+ (package-with-python2 python-autograd))
+
+(define-public lightgbm
+ (package
+ (name "lightgbm")
+ (version "2.0.12")
+ (source (origin
+ (method url-fetch)
+ (uri (string-append
+ "https://github.com/Microsoft/LightGBM/archive/v"
+ version ".tar.gz"))
+ (sha256
+ (base32
+ "132zf0yk0545mg72hyzxm102g3hpb6ixx9hnf8zd2k55gas6cjj1"))
+ (file-name (string-append name "-" version ".tar.gz"))))
+ (native-inputs
+ `(("python-pytest" ,python-pytest)
+ ("python-nose" ,python-nose)))
+ (inputs
+ `(("openmpi" ,openmpi)))
+ (propagated-inputs
+ `(("python-numpy" ,python-numpy)
+ ("python-scipy" ,python-scipy)))
+ (arguments
+ `(#:configure-flags
+ '("-DUSE_MPI=ON")
+ #:phases
+ (modify-phases %standard-phases
+ (replace 'check
+ (lambda* (#:key outputs #:allow-other-keys)
+ (with-directory-excursion ,(string-append "../LightGBM-" version)
+ (invoke "pytest" "tests/c_api_test/test_.py")))))))
+ (build-system cmake-build-system)
+ (home-page "https://github.com/Microsoft/LightGBM")
+ (synopsis "Gradient boosting framework based on decision tree algorithms")
+ (description "LightGBM is a gradient boosting framework that uses tree
+based learning algorithms. It is designed to be distributed and efficient with
+the following advantages:
+
+@itemize
+@item Faster training speed and higher efficiency
+@item Lower memory usage
+@item Better accuracy
+@item Parallel and GPU learning supported (not enabled in this package)
+@item Capable of handling large-scale data
+@end itemize\n")
+ (license license:expat)))
+
+(define-public vowpal-wabbit
+ ;; Language bindings not included.
+ (package
+ (name "vowpal-wabbit")
+ (version "8.5.0")
+ (source (origin
+ (method url-fetch)
+ (uri (string-append
+ "https://github.com/JohnLangford/vowpal_wabbit/archive/"
+ version ".tar.gz"))
+ (sha256
+ (base32
+ "0clp2kb7rk5sckhllxjr5a651awf4s8dgzg4659yh4hf5cqnf0gr"))
+ (file-name (string-append name "-" version ".tar.gz"))))
+ (inputs
+ `(("boost" ,boost)
+ ("zlib" ,zlib)))
+ (arguments
+ `(#:configure-flags
+ (list (string-append "--with-boost="
+ (assoc-ref %build-inputs "boost")))))
+ (build-system gnu-build-system)
+ (home-page "https://github.com/JohnLangford/vowpal_wabbit")
+ (synopsis "Fast machine learning library for online learning")
+ (description "Vowpal Wabbit is a machine learning system with techniques
+such as online, hashing, allreduce, reductions, learning2search, active, and
+interactive learning.")
+ (license license:bsd-3)))
+
+(define-public python2-fastlmm
+ (package
+ (name "python2-fastlmm")
+ (version "0.2.21")
+ (source
+ (origin
+ (method url-fetch)
+ (uri (pypi-uri "fastlmm" version ".zip"))
+ (sha256
+ (base32
+ "1q8c34rpmwkfy3r4d5172pzdkpfryj561897z9r3x22gq7813x1m"))))
+ (build-system python-build-system)
+ (arguments
+ `(#:python ,python-2)) ; only Python 2.7 is supported
+ (propagated-inputs
+ `(("python2-numpy" ,python2-numpy)
+ ("python2-scipy" ,python2-scipy)
+ ("python2-matplotlib" ,python2-matplotlib)
+ ("python2-pandas" ,python2-pandas)
+ ("python2-scikit-learn" ,python2-scikit-learn)
+ ("python2-pysnptools" ,python2-pysnptools)))
+ (native-inputs
+ `(("unzip" ,unzip)
+ ("python2-cython" ,python2-cython)
+ ("python2-mock" ,python2-mock)
+ ("python2-nose" ,python2-nose)))
+ (home-page "http://research.microsoft.com/en-us/um/redmond/projects/mscompbio/fastlmm/")
+ (synopsis "Perform genome-wide association studies on large data sets")
+ (description
+ "FaST-LMM, which stands for Factored Spectrally Transformed Linear Mixed
+Models, is a program for performing both single-SNP and SNP-set genome-wide
+association studies (GWAS) on extremely large data sets.")
+ (license license:asl2.0)))