add columnSimilarities from Spark MLlib RowMatrix

using computeCost() in Spark v1.5.0

fix download

working PySpark.reduceByKey() of hw2 q1

considering using Spark.reduceByKey() with customized function

Spark.groupByKey() took too much time/memory to run

draft of spark

draft of spark: to include original data in the map after filter

remove excess print

first working q1 purely by numpy and scipy, no map/reduce

scipy.spatial.distance try

add transposed toy data from 1st rs slides

add from sklearn.metrics.pairwise import cosine_similarity

homeworks from cs100.1x part 3

start q1 Recommendation Systems