zhmg23

我们是如此的不同

rhel7.2从头构建和安装xgboost包

xgboost 是"极端梯度上升"(Extreme Gradient Boosting)的简称, 它类似于梯度上升框架,但是更加高效。它兼具线性模型求解器和树学习算法。因此,它快速的秘诀在于算法在单机上也可以并行计算的能力。

这使得xgboost至少比现有的梯度上升实现有至少10倍的提升。它提供多种目标函数,包括回归,分类和排序。

下面是在rhel7.2从头构建和安装xgboost包的步骤:

安装前提:java、maven、git


设置java环境变量

# vi /etc/profile

export JAVA_HOME=/usr/local/jdk

export JRE_HOME=${JAVA_HOME}/jre

export CLASSPATH=.:${JAVA_HOME}/lib:${JRE_HOME}/lib

export PATH=${JAVA_HOME}/bin:$PATH

# source /etc/profile


#  yum install -y maven



# git clone --recursive https://github.com/dmlc/xgboost


# cd  xgboost

#  make -j4

#  cd jvm-packages/


#  mvn package

 如上,表示构建jar包完成。 



注:如果出现这个错误

[ERROR] Failed to execute goal org.scalastyle:scalastyle-maven-plugin:0.8.0:check (checkstyle) on project xgboost-jvm: Execution checkstyle of goal org.scalastyle:scalastyle-maven-plugin:0.8.0:check failed: Plugin org.scalastyle:scalastyle-maven-plugin:0.8.0 or one of its dependencies could not be resolved: Could not transfer artifact org.scala-lang:scala-library:jar:2.11.6 from/to central (https://repo.maven.apache.org/maven2): GET request of: org/scala-lang/scala-library/2.11.6/scala-library-2.11.6.jar from central failed: Premature end of Content-Length delimited message body (expected: 5564007; received: 901957 -> [Help 1]

  

屏蔽pom.xml 此文件中如下内容

            <!-- <plugin>

                <groupId>org.apache.maven.plugins</groupId>

                <artifactId>maven-checkstyle-plugin</artifactId>

                <version>2.17</version>

                <configuration>

                    <configLocation>checkstyle.xml</configLocation>

                    <failOnViolation>true</failOnViolation>

                </configuration>

                <executions>

                    <execution>

                        <id>checkstyle</id>

                        <phase>validate</phase>

                        <goals>

                            <goal>check</goal>

                        </goals>

                    </execution>

                </executions>

            </plugin> -->

然后在重新执行 mvn package



 参考:https://docs.databricks.com/user-guide/faq/xgboost.html

   https://xgboost.readthedocs.io/en/latest/jvm/#installation



评论