One of the first things a traveller notices when going to Costa Rica is that almost every town or city is named after a Saint. Growing up in Costa Rica, I just assumed this was something true of all Latin American countries, a legacy of Spanish colonial times. California has San Jose, San Fransisco, etc. I thought it was like that from California to Argentina.
After some travelling in Latin America I realised this was simply not true.
Graphs to accompany my poster presentation.
The data set The generated data is plotted below: Metropolis Hastings Algorithm This visual explanation of the Metropolis Hastings algorithm shows the proposal and the true density along with the Metropolis Hastings step. It was inspired and adapted from code by Balazs Torok.
Diagnostic Plots for Beta 2 via Gibbs Sampler Autocorrelation Plot The autocorrelations are negligable, suggesting good mixing.
Running Means Plot The running means have settled.
Three chain trace plot All three chains seem to be in the same place.
This week I had an interesting conversation (as interesting as these conversations can get) about the use of spreadsheets in business. I am still a little bit surprised by how often they are used for things they are not really good for. Now, I don’t want to write about how there are better alternatives to spreadsheets. Instead, I’d like to illustrate how powerful non-spreadsheet tools are by describing a task I completed this week using R.