Create 06_workshop_extra.R

Author: timothyfraser

code/06_workshop_extra.R

@@ -0,0 +1,83 @@
+# 06_workshop_extra.R
+
+
+# EXERCISE 1 - Writing a Function #####################3
+f = function(Eu, El, sd){     
+  cp = (Eu - El) / (6 * sd)  
+  return(cp)
+}
+
+# roxygen2 style commenting for functions
+
+#' @name f
+#' @title Cp Process Control Index Function
+#' @description 
+#' Calculate any Cp statistic with just 3 inputs! Yay!
+#' @param Eu:int Upper Specification Limit
+#' @param El:int Lower Specification Limit
+#' @param sd:int Sigma-short (average within group standard deviation)
+#' @importFrom dplyr `%>%` filter (for example)
+f = function(Eu, El, sd){     
+  cp = (Eu - El) / (6 * sd)  
+  return(cp)
+}
+
+
+f(Eu = 10, El = 2, sd = 5)
+
+# EXERCISE 2 ###################################
+library(dplyr)
+library(readr)
+water = read_csv("workshops/onsen.csv", show_col_types = FALSE)
+water %>% head() %>% glimpse()
+
+
+
+
+
+
+# EXERCISE 3 #####################################
+
+
+# What's bootstrapping?
+# Take the existing data
+# Sample with Replacement for the entire data
+# Calculate a Stat
+# Get a stat that has been slightly influenced by random chance
+
+
+
+
+library(dplyr)
+library(readr)
+water = read_csv("workshops/onsen.csv", show_col_types = FALSE)
+
+water %>%
+  summarize(sigma_t = sd(temp))
+
+
+# We need 1000 standard deviations
+# We need 1000 resampled water datasets
+# We need 1000 water datasets --- x
+# We need a dataframe of 1000 ids -- x
+
+boot = tibble(rep = 1:1000) %>%
+  group_by(rep) %>%
+  reframe(water) %>%
+  group_by(rep) %>%
+  sample_n(size = n(), replace = TRUE)
+
+stat = boot %>%
+  group_by(rep) %>%
+  summarize(sigma_t = sd(temp))
+
+stat$sigma_t %>% hist()
+
+stat %>% 
+  ungroup() %>%
+  summarize(
+    estimate = mean(sigma_t),
+    se = sd(sigma_t),
+    upper = estimate + se*qnorm(0.975),
+    lower = estimate - se*qnorm(0.975) 
+  )