This repo is an example of Hadoop Map-Combiner-Reduce operation over CSV files.
Regression is a Java program that gathers vehicle data from Vehicules.csv file, representing a vehicle sample.
The goal of this is to calculate a covariance between the weight and the consumption of the vehicle.
The Vehicules.csv is structured like this :
| Weight (kg) | Consumption (liters) |
|---|
With hadoop installed, you must put the file on the hadoop disk :
hadoop fs -put Vehicules.csv /test
Next, after having compiled the project (with Maven for example : mvn clean package), you will execute the project :
hadoop jar NameOfYourJar.jar RegressionDriver /test/Vehicules.csv /results
You can see the results using 
X = Weight
Y = Consumption
n 28.0
Sx 33515.0
Sx² 4.2694125E7
X Variance 92066.67729591834
Sy 254.1
Sy² 2440.5699999999993
Sum X x Y 321404.5
X average 1196.9642857142858
X² average 1524790.1785714286
Y average 9.075
Y² average 87.16321428571426
X x Y average 11478.732142857143
X variance 92066.67729591834
Y variance 4.807589285714272
Covariance 616.28125
Corrélation r 0.9263263981866983
β0 1.0626912269314541
β1 0.006693857844127085
yi = 1.0626912269314541 + 0.006693857844127085 * xi + ei 0.0
The correlation between the weight and the consumption of the vehicle is very good !