Mapping Historical Networks, pt.1

Here the nodes are separated by category of connection and each ‘edge’ or connection is labeled with a year. The same data can be organized multiple ways by sorting and filtering certain data sets.

This project is a spontaneous offshoot of the research that I have been doing for Michael Henry, which is focusing on establishing Thomas Jefferson’s knowledge of environmental architectural design throughout the conception, execution, and modification of Monticello.

It all began with the suggestion from Prof. Henry that it might be beneficial to create something like a “LinkedIn” diagram for Mr. Jefferson to begin to understand what he knew, where he might have learned it, and who he might have learned it from. I started this task manually using Illustrator. This initial diagram began as a circular chart of sorts, starting with a nexus being Thomas Jefferson and radiating outward by year, and divided into ‘hemispheres’–the top half being Europe, the bottom being the United States. This was utterly time consuming, and frustrating on the mind-numbing level to know that as research progressed I would have to redraw this chart. Each time. I found it incredibly hard to believe that there was no easier way, more specifically no parametric way that would streamline research and visualization simultaneously.

A long spring break later, I came across a way to begin organizing the data that I had already found in the form of citations, etc. and visualizing it. I turned to to the same methods and software that is being used to organize and visualize Big Data, such as social networks and google searches, to solve this.


I first found a project being undertaken by Harvard called Visualize Historical Networks which attempts much of the same thing using a software, I believe, called Gephi–which uses Java and other things I don’t yet understand to begin to visualize adjacency at a specific time and extended periods.

I came across a plugin for excel that allowed me to do this quickly and easily (as this in itself is not part of my research) using my Endnote database. It is currently limited in its scope, ability to modify parameters, and export options however not totally limited. I have already drafted a proposal to expand this research using some creative scripting to communicate between programs, MatLab, and a brute-force method for harvesting historical data from large digital databases (Library of Congress ,etc.).

Adjacency matrix from excel, which can then be used in MatLab to perform much more powerful computations. This is the data language of such networks.

We’ll see where this goes, time permitting!

Old bricks & New Tricks

I’m taking the opportunity to apply some new software (ScanAndSolve™) to begin to visualize some of the potential mechanics involved with a 19th century cast iron element atop a masonry structure that is part of a project for the Building Pathology class I am in this semester with Prof. Michael Henry.

It should be noted that this is only a tool, and even this entire portion of the project–that being the analysis of possible structural failure mechanisms–is but a very small piece of the puzzle. In fact, it merely represents the final stage; a thorough investigation must be first done of the system as a whole that takes into account the construction, material properties, material adjacency, maintenance history, weather data, to name but a few.

To that, SnS™ offers this disclaimer:

“Scan&Solve(tm) Results

Design decisions require experimental data and substantial experience; they should never be made based solely on a software simulation. Simulation is not intended to replace physical testing of prototypes, which is required to validate any design.”

To be very honest, I just wanted to try out this new toy after applying for the student license (which ScanAndSolve™ was quite happy to do, unlike many other companies that make the process difficult).

More to come!

The images below map the principle tension and compression forces in lb/ft^2.                     Malleable Cast Iron was used as the material (only 2 offered at the moment, the other which is nodular was not invented until 1943 and would not have been used here)

E+ 3.551e+09 lb/ft^2                                                                                                                 Sut= 7.10105e+06 lb/ft^2                                                                                                         rho= 1.430e+01 slug/ft^3

This slideshow requires JavaScript.

Here is a quick description of the software and its comparison to traditional FEA (finite element analysis:

“Scan&Solve™ for Rhino for Windows completely automates basic structural simulation of Rhino solids. Unlike other analysis tools, no preprocessing (meshing, simplification, healing, translating, etc.) is needed.”


Essentially its an incredibly fast, streamlined, and easy-to-apply software to be applied quickly, multiple times though a design process to visualize the system. It’s not perfect, no program is–even the stand-alone, multi-thousand dollar ones, aren’t–but I’ve found it to be quite helpful.

Anyway, go check it out for yourself.