Publications by Mickey Campbell
Visibility Rescaling Functions
The logistic function The general form of the logistic function is: \[f(x)=\frac{L}{1+e^{-k(x-x_{0})}}\] where \(x_{0}\) is the midpoint (or inflection point at which the rate of change switches from positive to negative) of the logistic curve, \(L\) is the maximum value at which the curve levels out (or its asymptote), and \(k\) is the steepn...
2501 sym R (2033 sym/7 pcs) 5 img
VI Paper Data Simulation Exercise
Introduction In our VI paper, there is an interesting “problem”, where site-level models are, on average, performing worse than national-level models. But, when the predictions vs. observations are aggregated among all site-level and national-level models, it gives the appearance that site-level models are better. Whenever faced with wei...
4563 sym 3 img
Mixed Effects RF Testing
Introduction I wanted to play around with the mixed effects random forest, and do some comparisons between a basic RF, and I thought it would be useful to share what I found, particularly for Jessie. For the sake of simplicity, I’m just going to do this analysis using the field data we collected (omitting the data compiled from a whole bunc...
6322 sym R (12463 sym/16 pcs) 4 img
PJ Species Distribution Modeling with FIA Data Preparation
INTRODUCTION The objective of this document is to try to gain a better understanding of what FIA data will be needed for a project aimed at mapping the distribution of piñon and juniper tree species in the US. # load libraries library(stringr) library(magrittr) library(data.table) library(sf) ## Linking to GEOS 3.9.1, GDAL 3.2.1, PROJ 7.2.1;...
6883 sym R (17962 sym/34 pcs) 6 img
Axis Confusion
First things first – I’m going to create two plotting functions, one for basic x-y scatterplots, and one for predicted vs. observed scatterplots, to minimize repetitive plotting code throughout this document. Please reserve all questions about my continued use of base R and my resistance against the Tidyverse cult until the end… Of tim...
3566 sym 3 img
Quantile-based bias correction
# simulate data x <- rnorm(1000) y <- x + rnorm(1000) # split into training/validation/test x.train <- x[1:600] y.train <- y[1:600] x.valid <- x[601:800] y.valid <- y[601:800] x.test <- x[801:1000] y.test <- y[801:1000] # generate linear model and predict on validation data mod <- lm(y.train ~ x.train) pred.valid <- predict(mod, lis...
11 sym 4 img