Python

Big data: Python for ENVI-met

Over the last years, we experience an increasing availability of huge datasets spanning a wide range of topics from personal life over image analysis up to specific scientific data. In synchronicity with the availability of such big data, the question arises how we can manage these gigabytes of information and extract and visualize the information we are looking for.Obviously, the answer to this question cannot be given by a single person, not even a single company even if it includes Google, Amazon or Microsoft.

It is almost impossible to guess in advance, which questions a user might want to ask to a data set in advance and then provide these analysis tools pre-coded in your piece of software. Hence, the analysis of big date requires powerful scripting engines and a wide universe of different libraries that can be used. In the last years, Python has established itself as the core scripting language for all kind of data analysis with libraries liking it to massive computing platforms such as TensorFlow.

ENVI-met simulations also create a huge amount of data and there are many perspectives on how we can look at these data. Although it has always been the core concept of ENVI-met to make simulations and visualisations as simple as possible, we will open our data to Python world step-by-step in the next releases.

The up-to-come ENVI-met TreePass was the role model where we experienced the unlimited possibilities of Python for data analysis and visualisation that goes far ahead of what practically can be implemented fixed in a software product.

Using Python for ENVI-met will allow to use any existing Python library to load, analyse and visualize simulation data of different kinds being it atmospheric profiles or soil data.

From the ENVI-met side, compiled objects will be exposed to Python allowing an easy and fast access to the ENVI-met data models including the use of the well-known navigation tools such as the datafile map from LEONARDO.