Posterior explanation

paper/paper.md

@@ -94,15 +94,14 @@ The demo problem allows users to build intuition on how $V_{prior}$ and $V_{perf
 
 This figure provides a visual example of the empirical likelihood (bars) of the electrical conductivity assosciated with 
 positive sites (green) and negative sites (red): 
-![Likelihoods of Electrical Conductivity of Lower Crust for INGENIOUS area.\label{fig:Likelihood_CondLowCrust}][width=0.7](Likelihood_CondLowCrust.png)
-
-\includegraphics[width=0.7\textwidth,height=\textheight]{Likelihood_CondLowCrust.png}
+![Likelihoods of Electrical Conductivity of Lower Crust for INGENIOUS area.\label{fig:Likelihood_CondLowCrust}](Likelihood_CondLowCrust.png)
 
 The continuous lines in \autoref{fig:Likelihood_CondLowCrust.png} are the fitted kde likelihoods of the optimal bandwidth given the grid search for accuracy in Naïve Bayes. The likelihood scaled to a prior probability of success ($Pr(\Theta = positive$)=0.2) 
 ![Prior-Scaled Likelihoods of Electrical Conductivity of Lower Crust for INGENIOUS area.\label{fig:Scaled20Likelihood_CondLowCrust}](Scaled20Likelihood_CondLowCrust.png){width=50%} 
 
 The posterior plot of $Pr(\Theta = positive$)=0.2 are shown below
-![Posterior of Electrical Conductivity of Lower Crust for INGENIOUS area.\label{fig:Posterior_Prior20}](Posterior_Prior20.png){ width=0.5} 
+![Posterior of Electrical Conductivity of Lower Crust for INGENIOUS area.\label{fig:Posterior_Prior20}](Posterior_Prior20.png)
+This plot shows, given any of the conductivity bin values are observed, how likely are you in a positive (green) versus negative (red) hidden geothermal resource. In general, electrical conductivities of lower crust less than 4.5 (S-m) are more likely to be a negative geothermal site, whereas greater than 4.75 S-m the calibrated dataset show $Pr(\Theta = positive)$= 100%. 
 
 # Acknowledgements
 We acknowledge contributions from Sierra Rosado during the genesis of this project and from Nicole Taverna for providing feedback.