A few more edits from review of text

sryza · May 23, 2017 · 1643bfd · 1643bfd
1 parent 1eb2e50
commit 1643bfd
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 7 deletions.
diff --git a/ch02-intro/src/main/scala/com/cloudera/datascience/intro/RunIntro.scala b/ch02-intro/src/main/scala/com/cloudera/datascience/intro/RunIntro.scala
@@ -33,16 +33,15 @@ object RunIntro extends Serializable {
 
     val preview = spark.read.csv("hdfs:///user/ds/linkage")
     preview.show()
-    preview.schema.foreach(println)
+    preview.printSchema()
 
     val parsed = spark.read
       .option("header", "true")
       .option("nullValue", "?")
       .option("inferSchema", "true")
       .csv("hdfs:///user/ds/linkage")
     parsed.show()
-    val schema = parsed.schema
-    schema.foreach(println)
+    parsed.printSchema()
 
     parsed.count()
     parsed.cache()
@@ -61,7 +60,7 @@ object RunIntro extends Serializable {
     summary.select("summary", "cmp_fname_c1", "cmp_fname_c2").show()
 
     val matches = parsed.where("is_match = true")
-    val misses = parsed.filter($"is_match" === lit(false))
+    val misses = parsed.filter($"is_match" === false)
     val matchSummary = matches.describe()
     val missSummary = misses.describe()
 
@@ -76,9 +75,9 @@ object RunIntro extends Serializable {
     """).show()
 
     val matchData = parsed.as[MatchData]
-    val scored = matchData.map(md => {
+    val scored = matchData.map { md =>
       (scoreMatchData(md), md.is_match)
-    }).toDF("score", "is_match")
+    }.toDF("score", "is_match")
     crossTabs(scored, 4.0).show()
   }
 

diff --git a/ch06-lsa/src/main/scala/com/cloudera/datascience/lsa/RunLSA.scala b/ch06-lsa/src/main/scala/com/cloudera/datascience/lsa/RunLSA.scala
@@ -39,7 +39,6 @@ object RunLSA {
     vecRdd.cache()
     val mat = new RowMatrix(vecRdd)
     val svd = mat.computeSVD(k, computeU=true)
-    val u = svd.U.rows.zipWithUniqueId()
 
     println("Singular values: " + svd.s)
     val topConceptTerms = topTermsInTopConcepts(svd, 10, 10, termIds)