- São Paulo
Block or Report
Block or report GuiMarthe
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePinned
-
PyDataSP_pandas_Ago2017
PyDataSP_pandas_Ago2017 PublicEste repo contem a minha palestra na PyData São Paulo sobre Análise de Dados + pandas.
-
A simple procedure for sampling a di...
A simple procedure for sampling a distribution to look like another. A method through binning and another by kde estimation. The binning idea came from this stats exchange question and the kde method came from other studies of mine. 1library(tidyverse)
2library(broom)
345df <-
-
Little function I created in R for a...
Little function I created in R for adding all lagged values up to n of a variable to a df. Can be improved for handling more than one variable. 1add_lagged <- function(df, var, n = 1) {
2var <- enquo(var)
3names <- map(1:n, ~ paste0(quo_name(var), '_lag_' ,.))
45lagged_cols <- map2(1:n, names, ~ df %>% transmute(!!.y := lag(!!var, n = .x))) %>%
-
A small little idea on implementing ...
A small little idea on implementing bootstrap only using purrr, not dplyr based, after reading a google data science blog 1## http://www.unofficialgoogledatascience.com/2015/08/an-introduction-to-poisson-bootstrap26.html
23n <- 10000000
4data <- rnorm(n, mean = 4, sd = 2)
5 -
quick log odds, odds, probability eq...
quick log odds, odds, probability equivalence 1library(dplyr)
234tibble(
5prob = seq(0, 1, 0.01),
-
Um simples script que cria 300 variá...
Um simples script que cria 300 variáveis normais de 100 pares de média e variância distintas e calcula suas médias e intervalos de confiança por meio do bootstrap e ainda calcula a proporção de vezes que a verdadeira média esta no intervalo 1library(tidyverse)
23### criando 300 váriáveis normais para 20 diferentes
4### pares de média e variância distintos
5
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.