H2O makes Hadoop do math! H2O scales statistics, machine learning and math over BigData. H2O is extensible and users can build blocks using simple math legos in the core. H2O keeps familiar interfaces like R, Python, Excel & JSON so that BigData enthusiasts & experts can explore, munge, model and score datasets using a range of simple to advanced algorithms. Data collection is easy. Decision making is hard. H2O makes it fast and easy to derive insights from your data through faster and better predictive modeling. H2O allows online scoring and modeling in a single platform.
- Downloading H2O Dev
- Open Source Resources
- Using H2O Dev Artifacts
- Building H2O Dev
- Launching H2O after Building
- Building H2O on Hadoop
- Sparkling Water
- Documentation
- Community / Advisors / Investors
While most of this README is oriented towards developers who do their own builds, most H2O users just download and use a pre-built version. If that's you, just follow these steps:
- Point to http://h2o.ai
- Click on Download
- Scroll down to find the section for H2O Dev
- Click on the version you want (generally the latest numbered release)
- Github
- JIRA (PUBDEV contains issues for the current H2O Dev project; PUB contains issues for the legacy H2O project)
- h2ostream community forum (this is where to ask a question)
- Documentation
- Bleeding edge nightly build page: http://s3.amazonaws.com/h2o-release/h2o-3/master/latest.html
- FAQ: http://h2o.ai/product/faq/
- Download (pre-built packages)
- Jenkins
- Website
- Follow us on Twitter, @h2oai
Every nightly build publishes R, Python, Java and Scala artifacts to a per-build repository. In particular, you can find java artifacts in the maven/repo directory.
Here is an example snippet of a gradle build file using h2o-3 as a dependency. Replace x, y, z, and nnnn with valid numbers.
// h2o-3 dependency information
def h2oBranch = 'master'
def h2oBuildNumber = 'nnnn'
def h2oProjectVersion = "x.y.z.${h2oBuildNumber}"
repositories {
// h2o-3 dependencies
maven {
url "https://s3.amazonaws.com/h2o-release/h2o-3/${h2oBranch}/${h2oBuildNumber}/maven/repo/"
}
}
dependencies {
compile "ai.h2o:h2o-core:${h2oProjectVersion}"
compile "ai.h2o:h2o-algos:${h2oProjectVersion}"
compile "ai.h2o:h2o-web:${h2oProjectVersion}"
compile "ai.h2o:h2o-app:${h2oProjectVersion}"
}
See the latest H2O Dev bleeding edge nightly build page for information about installing nightly build artifacts.
See the h2o-droplets github repository for a working example of how to use java artifacts with gradle.
Note: Stable H2O Dev artifacts are periodically published to Maven Central (click here to search) but may lag substantially behind H2O Dev Bleeding Edge nightly builds.
Getting started with H2O development requires JDK 1.7, Node.js, and Gradle. We use the Gradle wrapper (called gradlew) to ensure an up-to-date local version of Gradle and other dependencies are installed in your development directory.
The assumption is the setup steps described in section 5.2 and beyond have been followed.
# Build H2O
git clone https://github.com/h2oai/h2o-3.git
cd h2o-3
./gradlew build -x test
# Start H2O
java -jar build/h2o.jar
# Point browser to http://localhost:54321
git clone https://github.com/h2oai/h2o-3.git
cd h2o-3
./gradlew syncSmalldata
./gradlew build
Note: Running tests starts 5 test JVMs that form an H2O cluster, and requires at least 8GB of RAM. Preferably 16GB of RAM.
git pull
./gradlew syncSmalldata
./gradlew clean
./gradlew build
A 'clean' is recommended after each git pull.
Skip tests by putting '-x test' at the end the gradle build command line. Tests typically run for 7-10 minutes on a Macbook Pro laptop with 4 CPUs (8 hyperthreads) and 16 GB of RAM.
Syncing smalldata is not strictly required after each pull, but if tests fail due to missing data files then this is the first troubleshooting step to try. Syncing smalldata grabs data files from AWS S3 to the smalldata directory in your workspace. The sync is incremental. Do not check these files in. The smalldata directory is in .gitignore. If you do not run any tests, you do not need the smalldata directory.
`pip install grip`
`pip install tabulate`
`pip install wheel`
Step 1: Download and install WinPython.
From the command line, validate python is using the newly-installed package. Update the Environment variable with the WinPython path.
`pip install grip`
`pip install tabulate`
`pip install wheel`
Install Java 1.7 and add the appropriate directory C:\Program Files\Java\jdk1.7.0_65\bin with java.exe to PATH in Environment Variables. To make sure the command prompt is detecting the correct Java version, run:
javac -version
The CLASSPATH variable also needs to be set to the lib subfolder of the JDK:
CLASSPATH=/<path>/<to>/<jdk>/lib
Install Node.js and add the installed directory C:\Program Files\nodejs, which must include node.exe and npm.cmd to PATH if not already prepended.
To install these packages from within an R session, enter:
R> install.packages("RCurl")
R> install.packages("rjson")
R> install.packages("statmod")
R> install.packages(c("devtools", "roxygen2", "testthat"))
Install R and add the preferred bin\i386 or bin\x64 directory to your PATH.
Note: Acceptable versions of R are >= 2.13 && <= 3.0.0 && >= 3.1.1.
To manually install packages, download the releases of the following R packages: bitops, devtools, digest, Rcpp, RCurl, rjson, roxygen2, statmod, stringr, and testthat.
cd Downloads
R CMD INSTALL bitops_x.x-x.zip
R CMD INSTALL RCurl_x.xx-x.x.zip
R CMD INSTALL rjson_x.x.xx.zip
R CMD INSTALL statmod_x.x.xx.zip
R CMD INSTALL Rcpp_x.xx.x.zip
R CMD INSTALL digest_x.x.x.zip
R CMD INSTALL testthat_x.x.x.zip
R CMD INSTALL stringr_x.x.x.zip
R CMD INSTALL roxygen2_x.x.x.zip
R CMD INSTALL devtools_x.x.x.zip
Finally, install Rtools, which is a collection of command line tools to facilitate R development on Windows. NOTE: During Rtools installation, do not install Cygwin.dll.
Step 6. Install Cygwin
NOTE: During installation of Cygwin, deselect the Python packages to avoid a conflict with the Python.org package.
If Cygwin is already installed, remove the Python packages or ensure that Native Python is before Cygwin in the PATH variable.
Step 8. Git Clone h2o-3
If you don't already have a Git client, please install one. The default one can be found here http://git-scm.com/downloads . Make sure that during the install command prompt support is turned on.
Download and update h2o-3 source codes:
git clone https://github.com/h2oai/h2o-3
cd h2o-3
gradlew build
If you encounter errors run again with
--stacktracefor more instructions on missing dependencies.
If you don't have Homebrew, we recommend installing it. It makes package management for OS X easy.
Install Java 1.7. To make sure the command prompt is detecting the correct Java version, run:
javac -version
Using Homebrew:
brew install node
Otherwise, install from the NodeJS website.
Install R and add the bin directory to your PATH if not already included.
Install the following R packages: RCurl, rjson, statmod, devtools, roxygen2 and testthat.
cd Downloads
R CMD INSTALL bitops_x.x-x.tgz
R CMD INSTALL RCurl_x.xx-x.x.tgz
R CMD INSTALL rjson_x.x.xx.tgz
R CMD INSTALL statmod_x.x.xx.tgz
R CMD INSTALL Rcpp_x.xx.x.tgz
R CMD INSTALL digest_x.x.x.tgz
R CMD INSTALL testthat_x.x.x.tgz
R CMD INSTALL stringr_x.x.x.tgz
R CMD INSTALL roxygen2_x.x.x.tgz
R CMD INSTALL devtools_x.x.x.tgz
To install these packages from within an R session:
R> install.packages("RCurl")
R> install.packages("rjson")
R> install.packages("statmod")
R> install.packages(c("devtools", "roxygen2", "testthat"))
Step 4. Git Clone h2o-3
OS X should have with Git installed. To download and update h2o-3 source codes:
git clone https://github.com/h2oai/h2o-3
cd h2o-3
./gradlew build
If you encounter errors run again with
--stacktracefor more instructions on missing dependencies.
sudo apt-get install npm
sudo ln -s /usr/bin/nodejs /usr/bin/node
npm install -g bower
Install Java 1.7. Installation instructions can be found here JDK installation. To make sure the command prompt is detecting the correct Java version, run:
javac -version
Installation instructions can be found here R installation. Click “Download R for Linux”. Click “ubuntu”. Follow the given instructions.
To install the required packages, follow the same instructions as for OS X above.
Step 4. Git Clone h2o-3
If you don't already have a Git client:
sudo apt-get install git
Download and update h2o-3 source codes:
git clone https://github.com/h2oai/h2o-3
cd h2o-3
./gradlew build
If you encounter errors run again with
--stacktracefor more instructions on missing dependencies.
Make sure that you are not running as root, since
bowerwill reject such a run.
On Ubuntu 13.10, the default Node.js (v0.10.15) is sufficient, but the default npm (v1.2.18) is too old, so use a fresh install from the npm website:
sudo apt-get install node
sudo ln -s /usr/bin/nodejs /usr/bin/node
wget http://npmjs.org/install.sh
sudo apt-get install curl
sudo sh install.sh
For users of Intellij's IDEA, generate project files with:
./gradlew idea
For users of Eclipse, generate project files with:
./gradlew eclipse
java -jar build/h2o.jar
Pre-built H2O-on-Hadoop zip files are available on the download page. Each Hadoop distribution version has a separate zip file in h2o-3.
To build H2O with Hadoop support yourself, do the following from the top-level h2o-3 directory:
(export BUILD_HADOOP=1; ./gradlew build -x test)
./gradlew dist
This will create a directory called 'target' and generate zip files there. Note that BUILD_HADOOP is the default behavior when the username is 'jenkins' (see settings.gradle); otherwise you have to ask for it, as shown above.
In the h2o-hadoop directory each hadoop version has a build directory for the driver and an assembly directory for the fatjar.
You need to:
- Add a new driver directory and assembly directory (each with a build.gradle file) in h2o-hadoop
- Add these new projects to h2o-3/settings.gradle
- Add the new hadoop version to HADOOP_VERSIONS in make-dist.sh
- Add the new hadoop version to wget list in h2o-dist/index.html
These are the required steps to debug HDFS in IDEA as a standalone H2O process.
Debugging H2O on Hadoop as a hadoop jar hadoop mapreduce job is a non-trivial thing to do (read: I've never been able to do it and have no idea how; the obvious thing to try of attaching a remote debugger has never worked for me).
However, what you can do relatively easily is tweak the gradle settings for the project so that H2OApp has HDFS as a dependency. Here are the steps:
- Make the following changes to gradle build files below
- Change hadoop-client version in h2o-persist-hdfs to the desired version
- Add h2o-persist-hdfs as a dependency to h2o-app
- Close IDEA
- ./gradlew cleanIdea
- ./gradlew idea
- Re-open IDEA
- Run or debug H2OApp, and you will now be able to read from HDFS inside the IDE debugger
h2o-persist-hdfs is normally only a dependency of the assembly modules, since those are not used by any downstream modules. We want the final module to define its own version of hdfs if any is desired.
Note this example is for MapR 4, which requires the additional org.json dependency to work properly.
$ git diff
diff --git a/h2o-app/build.gradle b/h2o-app/build.gradle
index af3b929..097af85 100644
--- a/h2o-app/build.gradle
+++ b/h2o-app/build.gradle
@@ -8,5 +8,6 @@ dependencies {
compile project(":h2o-algos")
compile project(":h2o-core")
compile project(":h2o-genmodel")
+ compile project(":h2o-persist-hdfs")
}
diff --git a/h2o-persist-hdfs/build.gradle b/h2o-persist-hdfs/build.gradle
index 41b96b2..6368ea9 100644
--- a/h2o-persist-hdfs/build.gradle
+++ b/h2o-persist-hdfs/build.gradle
@@ -2,5 +2,6 @@ description = "H2O Persist HDFS"
dependencies {
compile project(":h2o-core")
- compile("org.apache.hadoop:hadoop-client:2.0.0-cdh4.3.0")
+ compile("org.apache.hadoop:hadoop-client:2.4.1-mapr-1408")
+ compile("org.json:org.json:chargebee-1.0")
}
Sparkling Water combines two open source technologies: Apache Spark and H2O, our machine learning engine. It makes H2O’s library of Advanced Algorithms, including Deep Learning, GLM, GBM, KMeans, PCA, and Random Forest, accessible from Spark workflows. Spark users are provided with options to select the best features from either platforms to meet their Machine Learning needs. Users can combine Spark's RDD API and Spark MLLib with H2O’s machine learning algorithms, or use H2O independent of Spark in the model building process and post-process the results in Spark.
Here are links to resources for Sparkling Water:
- Download page for pre-built packages (Scroll down for Sparkling Water)
- Sparkling Water github repository
- README
- Developer documentaton
To generate the REST API documentation, use the following commands:
cd ~/h2o-3
cd py
python ./generate_rest_api_docs.py # to generate Markdown only
python ./generate_rest_api_docs.py --generate_html --github_user GITHUB_USER --github_password GITHUB_PASSWORD # to generate Markdown and HTML
The default location for the generated documentation is build/docs/REST.
Documentation for each bleeding edge nightly build is reachable from the nightly build page.
We will breathe & sustain a vibrant community with the focus of taking software engineering approach to data science and empower everyone interested in data to be able to hack data using math and algorithms. Join us on google groups h2ostream.
Team & Committers
SriSatish Ambati
Cliff Click
Tom Kraljevic
Tomas Nykodym
Michal Malohlava
Kevin Normoyle
Spencer Aiello
Anqi Fu
Nidhi Mehta
Arno Candel
Josephine Wang
Amy Wang
Max Schloemer
Ray Peck
Prithvi Prabhu
Brandon Hill
Jeff Gambera
Ariel Rao
Viraj Parmar
Kendall Harris
Anand Avati
Jessica Lanford
Alex Tellez
Allison Washburn
Amy Wang
Erik Eckstrand
James Dean
Neeraja Madabhushi
Sebastian Vidrio
Ben Sabrin
Matt Dowle
Mark Landry
Tony D'Amato
Erin LeDell
Scientific Advisory Council
Stephen Boyd
Rob Tibshirani
Trevor Hastie
Systems, Data, FileSystems and Hadoop
Doug Lea
Chris Pouliot
Dhruba Borthakur
Jishnu Bhattacharjee, Nexus Venture Partners
Anand Babu Periasamy
Anand Rajaraman
Ash Bhardwaj
Rakesh Mathur
Michael Marks
Egbert Bierman
Rajesh Ambati