@Test
public void testFrequentItems() {
DataFrame df = context.table("testData2");
String[] cols = {"a"};
DataFrame results = df.stat().freqItems(cols, 0.2);
Assert.assertTrue(results.collect()[0].getSeq(0).contains(1));
}
@Ignore
public void testShow() {
// This test case is intended ignored, but to make sure it compiles correctly
DataFrame df = context.table("testData");
df.show();
df.show(1000);
}
/** See SPARK-5904. Abstract vararg methods defined in Scala do not work in Java. */
@Test
public void testVarargMethods() {
DataFrame df = context.table("testData");
df.toDF("key1", "value1");
df.select("key", "value");
df.select(col("key"), col("value"));
df.selectExpr("key", "value + 1");
df.sort("key", "value");
df.sort(col("key"), col("value"));
df.orderBy("key", "value");
df.orderBy(col("key"), col("value"));
df.groupBy("key", "value").agg(col("key"), col("value"), sum("value"));
df.groupBy(col("key"), col("value")).agg(col("key"), col("value"), sum("value"));
df.agg(first("key"), sum("value"));
df.groupBy().avg("key");
df.groupBy().mean("key");
df.groupBy().max("key");
df.groupBy().min("key");
df.groupBy().stddev("key");
df.groupBy().sum("key");
// Varargs in column expressions
df.groupBy().agg(countDistinct("key", "value"));
df.groupBy().agg(countDistinct(col("key"), col("value")));
df.select(coalesce(col("key")));
// Varargs with mathfunctions
DataFrame df2 = context.table("testData2");
df2.select(exp("a"), exp("b"));
df2.select(exp(log("a")));
df2.select(pow("a", "a"), pow("b", 2.0));
df2.select(pow(col("a"), col("b")), exp("b"));
df2.select(sin("a"), acos("b"));
df2.select(rand(), acos("b"));
df2.select(col("*"), randn(5L));
}
@Test
public void pivot() {
DataFrame df = context.table("courseSales");
Row[] actual =
df.groupBy("year")
.pivot("course", Arrays.<Object>asList("dotNET", "Java"))
.agg(sum("earnings"))
.orderBy("year")
.collect();
Assert.assertEquals(2012, actual[0].getInt(0));
Assert.assertEquals(15000.0, actual[0].getDouble(1), 0.01);
Assert.assertEquals(20000.0, actual[0].getDouble(2), 0.01);
Assert.assertEquals(2013, actual[1].getInt(0));
Assert.assertEquals(48000.0, actual[1].getDouble(1), 0.01);
Assert.assertEquals(30000.0, actual[1].getDouble(2), 0.01);
}
@Test
public void testCrosstab() {
DataFrame df = context.table("testData2");
DataFrame crosstab = df.stat().crosstab("a", "b");
String[] columnNames = crosstab.schema().fieldNames();
Assert.assertEquals("a_b", columnNames[0]);
Assert.assertEquals("2", columnNames[1]);
Assert.assertEquals("1", columnNames[2]);
Row[] rows = crosstab.collect();
Arrays.sort(rows, crosstabRowComparator);
Integer count = 1;
for (Row row : rows) {
Assert.assertEquals(row.get(0).toString(), count.toString());
Assert.assertEquals(1L, row.getLong(1));
Assert.assertEquals(1L, row.getLong(2));
count++;
}
}
@Test
public void testCollectAndTake() {
DataFrame df = context.table("testData").filter("key = 1 or key = 2 or key = 3");
Assert.assertEquals(3, df.select("key").collectAsList().size());
Assert.assertEquals(2, df.select("key").takeAsList(2).size());
}
@Test
public void testExecution() {
DataFrame df = context.table("testData").filter("key = 1");
Assert.assertEquals(1, df.select("key").collect()[0].get(0));
}
@Test
public void testCovariance() {
DataFrame df = context.table("testData2");
Double result = df.stat().cov("a", "b");
Assert.assertTrue(Math.abs(result) < 1.0e-6);
}
@Test
public void testCorrelation() {
DataFrame df = context.table("testData2");
Double pearsonCorr = df.stat().corr("a", "b", "pearson");
Assert.assertTrue(Math.abs(pearsonCorr) < 1.0e-6);
}
