foe_2013

5 Arrays and matrices

10 Writing your own functions

11 Statistical models in R

R is an integrated suite of software facilities for data manipulation, calculation and graphical display.

ANALYTICS

GRAPHICS AND VISUALIZATION

R APPLICATIONS and EXTENSIONS***

PROGRAMMING LANGUAGE FEATURES

Too Many Commands

Sloppy Control of Variables

Loop-a-phobia

Functions That Act Like Procedures

Naming and Renaming Variables is Way Too Complicated

Inability to Analyze Multiple Variables

Poor Ability to Select Variable Sets

Too Much Complexity

Lack of Graphical User Interface (GUI)

Time series data are widely seen in analytics. Some examples are stock indexes/prices, currency exchange rates and electrocardiogram (ECG)

Time Series Decomposition

Time Series Forecasting

Time Series Clustering

Time Series Classification

The preceding scatterplot demonstrates that these data may not be linear.

In this tutorial, we will explore the a linear, quadratic, and cubic model.

To center a variable, simply subtract its mean from each data point and save the result into a new R variable, as demonstrated below.

Once the input variable has been centered, the higher order terms can be created.

we can simply multiply our centered term from step one and save the result into a new R variable, as demonstrated below.

Now we have all of the pieces necessary to assemble our linear and curvilinear models.

In this case, the quadratic and cubic terms are not statistically significant themselves nor are their models statistically significant beyond the linear model.

it can be helpful to summarize and compare our potential models using the summary(MODEL) and anova(MODEL1, MODEL2,… MODELi) functions.

The regression topics covered in these tutorials can be mixed and matched to create exceedingly complex models.

The good news is that more complex models can be created using the same techniques covered here. The basic principles remain the same.

Un par de buenas noticias para finalizar este breve análisis. La primera es que disponemos de una suite llamada RStudio que hace algo más sencillo el trabajo con R. La segunda es que podemos aprender a usar R de forma bastante sencilla y gamificada a través de Try R, un curso online elaborado por la editorial O'Reilly usando la plataforma Code School de forma totalmente gratuita.

al margen de los datos estructurados que podemos extraer: usuario, conexiones con otros usuarios, fecha y hora de publicación

con el lenguaje R y las funciones especializadas podemos trasladar las informaciones

Una vez que hemos obtenido los datos en bruto

en tablas normalizadas que nos permitirán realizar una exploración analítica de los datos y su representación gráfica

el aspecto más interesante es poder extraer información significativa del propio texto

Asimismo, tenemos la posibilidad de realizar análisis, en el texto

, de actitudes positivas o negativas hacia un determinado acontecimiento, producto o servicio

Esta pensado para poder trabajar con muchos datos, es ideal para el trabajo en bigdata. Asimismo, no solo es capaz de tratar los datos sino que también es capaz de tomar esos datos, de tal forma que podemos automatizar el proceso de toma de datos (útil si esos datos se encuentran en internet y son elaborados por terceros).

Al tratarse de un software pensado en clave estadística, dispone de una mayor varidedad de gráficos

Hemos dicho que una de las peculiaridades de este software es que no dispone de una interfaz visual, que se utilizar a través de un lenguage de programación propio. Esta misma desventaja es en realidad una ventaja, ya que ese mismo hecho hace que tengamos más libertad para hacer lo que queramos

Just one simple line of [R] code that would work just as well for a 100-by-100 loss triangle as it would for a 10-by-10 triangle. No hidden cells in the spreadsheet, no risk of spreadsheet error. It's a little bit of code you could look at in one screen, it's replicable ... and this does all the work that a spreadsheet would do. 

He uses the Allstate Claim Prediction Challenge data (from a recent Kaggle competition) to fit a Tweedie model to 13 million records of claim data. (The Tweedie distribution is often used to model insurance claims, where many claims are exactly zero, and non-zero claims follow a continuous Gamma-like distribution.) Using the forthcoming rxGLM function, he fit the model to this large data set in just over two minutes (140.22 seconds) using a single quad-core PC.