在Mahout中实现Apriori算法的步骤如下：
1. 导入必要的库和函数：
```java
import org.apache.mahout.fpm.pfpgrowth.fpgrowth2.FPGrowth;
import org.apache.mahout.fpm.pfpgrowth.fpgrowth2.FPGrowthItemsets;
import org.apache.mahout.fpm.pfpgrowth.fpgrowth2.FPGrowthJob;
import org.apache.mahout.fpm.pfpgrowth.fpgrowth2.FPGrowthDriver;
```
2. 创建一个FPGrowth对象并设置参数：
```java
FPGrowth fpGrowth = new FPGrowth();
fpGrowth.setMinSupport(0.5);
fpGrowth.setNumGroups(50);
```
3. 读取数据集并进行格式转换：
```java
FPGrowthDriver.runFPGrowth(args, fpGrowth);
```
4. 运行Apriori算法并获取频繁项集：
```java
FPGrowthJob fpGrowthJob = new FPGrowthJob();
FPGrowthItemsets itemsets = fpGrowthJob.findFrequentItemsets(data, fpGrowth, true, false);
```
5. 输出频繁项集：
```java
for (FPGrowthItem item : itemsets.all()) {
System.out.println(item);
}
```
通过以上步骤，就可以在Mahout中实现Apriori算法并获取频繁项集。需要注意的是，在实际应用中，还需要根据具体数据集和需求调整参数和设置。