pachevalier’s gists

pachevalier / gmm_vs_lm.R

Created July 30, 2013 10:48

This gist compare the results of the gmm() and the lm() function in R using simulated data.

	library("gmm")
	set.seed(1234567)
	N <- 1000
	dd <- data.frame(id = 1:N)
	dd$u <- rnorm(N)
	dd$x <- 1 + rnorm(N)
	dd$y <- 1 + dd$x + dd$u
	m1 <- lm(y ~ x, data = dd)
	m2 <- gmm(y ~ x, x = ~ x, wmatrix = "ident", data = dd)
	coefficients(m1)

pachevalier / chow_test.R

Created August 1, 2013 14:11

	mc <- lm(formula = SBR ~ Age, data = ch)
	m1 <- lm(formula = SBR ~ Age, data = subset(ch, Sex == "M"))
	m2 <- lm(formula = SBR ~ Age, data = subset(ch, Sex == "F"))
	sc <- sum(mc$residuals^2)
	s1 <- sum(m1$residuals^2)
	s2 <- sum(m2$residuals^2)
	k <- 2
	# Test statistic
	fstat <- (sc - (s1 + s2)) / k / (s1 + s2) * (length(mc$residuals) - 2*k)
	fstat

pachevalier / piecewise_linear_reg.R

Last active December 9, 2016 12:57

	library("strucchange")
	library("lubridate")

	set.seed(1234567)
	N <- 60
	df <- data.frame(id = 1:N)
	df$date <- seq(as.Date("2013-07-01"), by = "day", along = df$id)
	df$date2 <- difftime(df$date, ymd("2013-07-01"), units = "day")
	df$date3 <- difftime(df$date, ymd("2013-08-01"), units = "day")
	difftime(ymd("2013-08-01"), ymd("2013-07-01"), units = "day")

pachevalier / segmented.R

Created August 1, 2013 17:06

	library("lubridate")
	library("strucchange")
	library("segmented")

	set.seed(1234567)
	N <- 60
	df <- data.frame(id = 1:N)
	df$date <- seq(as.Date("2013-07-01"), by = "day", along = df$id)
	df$date2 <- difftime(df$date, ymd("2013-07-01"), units = "day")
	df$date3 <- ifelse(df$date > as.Date("2013-08-01"), difftime(df$date, ymd("2013-08-01"), units = "day"), 0)

pachevalier / fake_line_plot.R

Created October 29, 2013 16:09

Fake Line PLot

	N <- 100
	fk <- data.frame(id = 1:N)
	fk$x <- 1 + rnorm(N)
	fk$y1 <- 1 + fk$x + rnorm(N)
	fk$y2 <- 2 + fk$x + rnorm(N)

	ggplot(data = fk, aes(x = x, y = y1)) +
	geom_line(color = "#339966", size = 2) +
	geom_line(aes(y = y2), color = "#990000", size = 2) +
	theme_tufte()

pachevalier / function_categorie.R

Created January 24, 2014 14:39

Cette fonction crée des catégories à partir d'une variable continue.

	categorie <- function(x, breaks) {
	c <- length(breaks)
	temp <- 0
	temp[x <= breaks[1]] <- 1
	for (k in 1:c-1) {
	temp[x > breaks[k] & x <= breaks[k+1]] <- k+1
	}
	temp[x > breaks[c]] <- c + 1
	return(temp)
	}

pachevalier / codeiso.csv

Last active January 4, 2016 19:09

Code ISO des pays au format CSV. Importé de http://en.wikipedia.org/wiki/ISO_3166-1_alpha-3

	ISO	country
	ABW	Aruba
	AFG	Afghanistan
	AGO	Angola
	AIA	Anguilla
	ALA	Åland Islands
	ALB	Albania
	AND	Andorra
	ARE	United Arab Emirates
	ARG	Argentina

pachevalier / code_ioc.csv

Created January 28, 2014 13:32

Le comité international olympique a un code pays spécifique différent des code ISO usuels. Ce fichier de correspondance a été extrait du site du comité international olympique par une méthode de scraping.

	code	country
	AFG	Afghanistan
	RSA	South Africa
	ALB	Albania
	ALG	Algeria
	GER	Germany
	AND	Andorra
	ANG	Angola
	ANT	Antigua and Barbuda
	AHO	Netherlands Antilles

pachevalier / arrange.R

Last active August 29, 2015 13:57

Problem with the output of arrange

	library("dplyr")
	set.seed(123)
	N <- 100
	df <- data.frame(id = 1:N, x = rnorm(N))
	df$x[runif(N) < .1] <- NA
	table(is.na(tdf$x))
	tdf <- tbl_df(df)
	out <- arrange(tdf, desc(x))
	out2 <- tdf[order(tdf$x, decreasing = TRUE),]

pachevalier / sample.R

Created May 7, 2014 17:05

	library("ggplot2")
	set.seed(1234)
	N <- 100
	df <- data.frame(i = 1:N, x = rnorm(N))
	df$y <- 1 + df$x + rnorm(N)
	df$z <- (runif(N) < .3)

	pdf("output/test.pdf")
	ggplot(data = df, aes(x = x, y = y, shape = z)) +
	geom_point(size = 3) +

Paul-Antoine pachevalier