Last week we tried to install Oracle R Distribution (ORD) in Oracle Linux 7.1 using Yum, which is the installation method recommended by Oracle. After following closely the instructions provided in the documentation, instead of the Oracle R Distribution 3.2.0, we found ourselves with the latest (3.2.3) version of GNU R installed. What had happened is that in our /etc/yum.repos.d, …
Limitations of Spark MLlib linear algebra module
A couple of days ago I stumbled upon some unexpected behavior of Spark MLlib (v. 1.5.2), while trying some ultra-simple operations on vectors. Consider the following Pyspark snippet: Clearly, what happens is that the unary operator – (minus) for vectors fails, giving errors for expressions like -x and -y+x, although x-y behaves as expected. The result of the last operation, …
How NOT to perform feature selection!
Cross-validation (CV) is nowadays being widely used for model assessment in predictive analytics tasks; nevertheless, cases where it is incorrectly applied are not uncommon, especially when the predictive model building includes a feature selection stage. I was reminded of such a situation while reading this recent Revolution Analytics blog post, where CV is used to assess both the feature selection …
Oracle R Enterprise 1.4: ore.make.names does not work for Oracle DB connections
I have reported in the past about some unexpected behavior issues of Oracle R Enterprise 1.4 ore.make.names function; nevertheless, back then I had only tried it with Hive connections. I tried to use it today with an Oracle database connection, and it doesn’t seem to work. Here is a reproducible example in Oracle Big Data Lite VM 4.2.1, using the …
Manipulating Hive tables with Oracle R connectors for Hadoop
In this post, we’ll have a look at how easy it is to manipulate Hive tables using Oracle R connectors for Hadoop (ORCH, presently known as Oracle R Advanced Analytics for Hadoop – ORAAH). We will use the weblog data from Athens Datathon 2015, which we have already loaded in a Hive table named weblogs, as described in more detail …
Augmenting PCA functionality in Spark 1.5
Surprisingly enough, although the relatively new Spark ML library (not to be confused with Spark MLlib) includes a method for principal components analysis (PCA), there is no way to extract some very useful information regarding the PCA transformation, namely the resulting eigenvalues (check the Python API documentation); and, without the eigenvalues, one cannot compute the proportion of variance explained (PVE), …
Log files exploration with Oracle Big Data Discovery 1.1
In a previous post, we described how we performed exploratory data analysis (EDA) in real-world log files, as provided by Skroutz.gr, the leading online company in Greece for online price comparison, in the context of Athens Datathon 2015. In the present post we will have a look at the same job as performed with Oracle Big Data Discovery (v. 1.1), …
Installing RStudio & additional R packages in Oracle Big Data Lite VM 4.2.1
I was very happy to find out that, in the latest version (4.2.1) of Oracle Big Data Lite VM, all the R-related issues I had located and reported in the past (see here and here) have been resolved. Nevertheless, some new issues have emerged. Below are my findings and workarounds (if you are in a hurry, feel free to jump …
Athens Datathon 2015: exploratory data analysis for anomaly detection & data quality
Together with my friend and former colleague Georgios Kaiafas, we formed a team to participate to the Athens Datathon 2015, organized by ThinkBiz on October 3; the datathon took place at the premises of Skroutz.gr, which was also the major sponsor and the data provider. It was the second such event organized in Athens, and you can see the Datathon …
Big Data Discovery configuration in Oracle Big Data Lite VM 4.2.1
The latest version (4.2.1) of Oracle Big Data Lite VM, among many additions, now includes also the much-expected Oracle Big Data Discovery (v. 1.1), which I had not played with so far (it is a new product); so I thought to take it for a ride. Since my test data included geolocation attributes (latitude/longitude), one of the first things I …