Innovating for you
At Active Analytics it is our passion to create products and services that add value to your efforts and activities.
Active Analytics products are listed below, our product development focuses on tools for analysis that will be useful for a range of applications:
Actuarial Reserving in the Julia Programming Language.
The Julia programming language is taking the world of data science by storm and Active Analytics is embracing this new technology. we have released an open source actuarial reserving package written in Julia. The package is under constant development and new features will be added with time. The package can be downloaded from our GitHub page. Our focus in this package is both the quality and the performance and we have achieved both.
- The Chain Ladder Algorithm.
Big data storage and access.
ActiveH5 is an R package for big data storage and access, it allows you to store and access large data frames and matrices both on file and in memory and data is stored in chunks. It uses the HDF5 file format on disk allowing fast I/O speed. In memory chunks of data are accessed by pointers allowing large data sets to be accessed & processed in R efficiently.
- You can store R's data frames objects either on disk as h5DF or in memory as h5MEMDF which exisit in R as reference class objects. On disk the h5DF object is stored in HDF5 format.
- The h5DF & h5MEMDF objects allow you to store R's numeric and factor data types in columns. Character data is automatically transformed to factor data.
- You can also store numeric matrices using the ActiveH5 package. On disk h5MAT and in memory h5MEMMAT both data types are expressed in R as reference class interfaces.
- You can append data to exisiting ActiveH5 objects for both the data frame and matrix objects and management of factor levels is taken care of by the package.
ActiveReg is a sibling R package to ActiveH5. It allows you to analyse big data using linear regression and generalized linear regression models.
The ActiveReg package allows data you have stored using the ActiveH5 package to be analysed using generalized linear models (GLM). It is designed to give you a similar access to GLM functionality that you currently enjoy with the lm() and glm() functions in R but to address this kind of analysis with large data sets.
- The ActiveReg R package allows you to run regression models on large data sets created with the ActiveH5 package.
- The functions h5lm() and h5glm() provide you with access to linear and generalized linear models.
- Provides you with a familiar interface to create models expressed using R's formula objects.
- Standard link functions R programmers are familiar with are supported: logit, probit, cauchit, cloglog, identity, log, sqrt, 1/mu^2, and inverse.
- Standard distributions are supported: Binomial, Poisson, Gaussain, Quasi-Poisson, Quasi-Binomial, Gamma, Inverse Gaussian, and Quasi.