Author: dlyubimov
Date: Sat Jun 13 00:38:16 2015
New Revision: 1685198
URL: http://svn.apache.org/r1685198
Log:
CMS commit to mahout by dlyubimov
Modified:
mahout/site/mahout_cms/trunk/content/users/sparkbindings/play-with-shell.mdtext
Modified:
mahout/site/mahout_cms/trunk/content/users/sparkbindings/play-with-shell.mdtext
URL:
http://svn.apache.org/viewvc/mahout/site/mahout_cms/trunk/content/users/sparkbindings/play-with-shell.mdtext?rev=1685198&r1=1685197&r2=1685198&view=diff
==============================================================================
---
mahout/site/mahout_cms/trunk/content/users/sparkbindings/play-with-shell.mdtext
(original)
+++
mahout/site/mahout_cms/trunk/content/users/sparkbindings/play-with-shell.mdtext
Sat Jun 13 00:38:16 2015
@@ -2,6 +2,8 @@
This tutorial will show you how to play with Mahout's scala DSL for linear
algebra and its Spark shell. **Please keep in mind that this code is still in a
very early experimental stage**.
+_(Edited for 0.10.2)_
+
## Intro
We'll use an excerpt of a publicly available [dataset about
cereals](http://lib.stat.cmu.edu/DASL/Datafiles/Cereals.html). The dataset
tells the protein, fat, carbohydrate and sugars (in milligrams) contained in a
set of cereals, as well as a customer rating for the cereals. Our aim for this
example is to fit a linear model which infers the customer rating from the
ingredients.
@@ -161,25 +163,12 @@ right angle. An easy way to add such a b
column of ones to the feature matrix `\(\mathbf{X}\)`.
The corresponding weight in the parameter vector will then be the bias term.
-Mahout's DSL offers a ```mapBlock()``` method for custom modifications of a
DRM. All the rows in a partition are merged to a block of the matrix which is
given to custom code in a closure. For our example, we invoke ```mapBlock```
with ```ncol = drmX.ncol + 1``` to let the system know that change the number
of columns of the matrix. The input to our closure is a ```block``` of the DRM
and an array of ```keys``` for the rows contained in the block. In order to add
a column, we first create a new block with an additional column, then copy the
data from the current block into the new block and finally set the last column
to ones and return the new block.
+Here is how we add a bias column:
<div class="codehilite"><pre>
-val drmXwithBiasColumn = drmX.mapBlock(ncol = drmX.ncol + 1) {
- case(keys, block) =>
- // create a new block with an additional column
- val blockWithBiasColumn = block.like(block.nrow, block.ncol + 1)
- // copy data from current block into the new block
- blockWithBiasColumn(::, 0 until block.ncol) := block
- // last column consists of ones
- blockWithBiasColumn(::, block.ncol) := 1
-
- keys -> blockWithBiasColumn
-}
+val drmXwithBiasColumn = drmX cbind 1
</pre></div>
-(This looks like a lot of work for something that would be simply `cbind(drmX,
1)` in R. Matrix-scalar
-`cbind` combination is still a TODO in Mahout's dialect, although `cbind`
exists for other operand type combinations.)
-
Now we can give the newly created DRM ```drmXwithBiasColumn``` to our model
fitting method ```ols``` and see how well the resulting model fits the training
data with ```goodnessOfFit```. You should see a large improvement in the result.
<div class="codehilite"><pre>
