emhart

### Get your climate data..

usmex <- c(273:284,328:365)
usmex.basin <- create_map_df(usmex)
temp.dat <- get_historical_temp(usmex, "decade" )
temp.dat <- subset(temp.dat,temp.dat$year == 2000 )
usmex.map.df <- climate_map(usmex.basin,temp.dat,return_map=F)

emhart

search_datatype<- function(datastype,term){
   results <- noaa_datatypes(dataset=datatype)
   s_parse <- function(x){
     if(sum(grep(term,x[2],ignore.case=TRUE)) > 0){
       return(c(x[1],x[2]))
     }
   }
   out <- unlist(lapply(results$dataTypeCollection$dataType,s_parse))
   out <- data.frame(matrix(out,ncol=2,nrow=(length(out)/2),byrow=T))
   colnames(out) <- c("ID","Description")

emhart

#### Doesn't work...
map <- map_data('world')

ggplot(data = map, aes(long, lat, group=group)) + geom_path() +coord_map(xlim=c(-170, -40), ylim=c(20, 83))
### Does work

test <- map
test <- subset(test,test$long > -170 & test$long < -40)
test <- subset(test,test$lat > 20 & test$lat < 83)
ggplot(data = test, aes(long, lat, group=group)) + geom_path() 

emhart

require(twitteR)
require(ggplot2)
library(ROAuth)
library(plyr)

### See this post on how to get your key and secret to put 
### in below as strings: https://dev.twitter.com/discussions/631
reqURL <- "https://api.twitter.com/oauth/request_token"
accessURL <- "http://api.twitter.com/oauth/access_token"
authURL <- "http://api.twitter.com/oauth/authorize"

emhart

### NLS multiple starting points search 

### formula: a standard R formula that you want to include in the nls() call

### inits: is a list of 2x1 vectors giving the upper and lower bounds of a uniform distribution to pull random starting values from
### The list should have names that correspond to the names of your parameters, similar to the way you would call them in nls()
### e.g. a good starting list would be: starts <- list(a=c(0,50),b=c(0,50))


### miter The maximum time to try and find a solution

emhart

#!/usr/bin/python

import requests
import re
import sys
import os


def download_file(url,path):
    if path[-1] == "/":

emhart

## Extract % variance explained from a multiple regression model for each parameter
## Inputs: a fitted model object
## Details:  Uses a simple formula for each parameter as SSparam / SST
## Returns: a p x 2 dataframe with parameters as rows and name and % explained are the two columns

per_exp <- function(mod){
  ## Get sums of squares total (SSt)
  squares <- anova(mod)
  sst <- sum(squares$Sum)
  var_exp <- vector()

emhart

# Clean up old installs, removing old RMendeley and old ROAuth

install.packages("devtools")
library(devtools)
install_github("ROAuth","duncantl")
install_github("RMendeley","ropensci")
mc <- mendeley_auth()

emhart

function (d, colClasses) 
{
    colClasses <- rep(colClasses, len = length(d))
    d[] <- lapply(seq_along(d), function(i) switch(colClasses[i], 
        numeric = as.numeric(d[[i]]), character = as.character(d[[i]]), 
        Date = as.Date(d[[i]], origin = "1970-01-01"), POSIXct = as.POSIXct(d[[i]], 
            origin = "1970-01-01"), factor = as.factor(d[[i]]), 
        as(d[[i]], colClasses[i])))
    d
}

emhart

### "Smart brute force" algorithm.  Works in most mortal situations.
### Works by looping through a matrix that is m by k and putting numbers in 
### Is aware of the top and left cell, and works by sampling based on the probabilities of 
### numbers that are left in the set of valid possible numbers, e.g. those not excluded by the neighborhood rules (von Neumann)
### Parameters:
### k -- The number of rows
### m -- The number of columns
### n -- 1:N vector of desired outcomes (plants in the blog post), e.g. if N = 3, your n parameter would be n <- 1:3
### maxiter -- The maximum number of iterations you want to use in the algorithm (default is 100.