In IsaacVerm/reproseries: example for the series of reproducible analysis posts on isaacverm.github.io

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(reproseries)
library(magrittr)

The reproseries package itself is not sufficient, we need magrittr to pipe functions together.

Introduction

This vignette shows what you can do with the functions in the reproseries package. Code is not hidden so you can follow along nicely.

Data

The functions are written to be used on the built-in nottem dataset. This dataset looks like this:

temperatures <- nottem %>%
  temperatureSeriesToDf

temperatures %>% str

We first transform nottem from a time series in a dataframe object. This way str gives us a clearer overview of what the dataset looks like.

Analysis phases

Tidy

The raw nottem dataset is not tidy since variables are spread across columns. There are 3 variables (month, year and temperature) but there are 12 columns (one for each month). Let's fix this:

temperatures <- temperatures %>% 
  tidyTemperatures

temperatures %>% str

Now we have the 3 variables we need. Month and year are clear, value refers to the temperature in fahrenheit. In the next step we'll deal with this vague value variable.

Transform

We're no Americans so transform fahrenheits to celsius.

temperatures <- temperatures %>% 
  addCelsiusColumn

temperatures %>% str

In preparation of the modeling and visualisation we already calculate an average temperature per year.

avg_temperature_by_year <- temperatures %>% 
  summariseAvgTemperatureByYear

avg_temperature_by_year %>% str

Visualize

How did the average temperature evolve over the years?

avg_temperature_by_year %>%
  plotAvgTemperaturesByYear

Model

Is there a trend in the average temperatures?

avg_temperature_by_year %>%
  calculateAutoCorrelation

0.10 is rather low so there doesn't seem to be any trend.

Summary

All the functions available in the reproseries package have been introduced above.

IsaacVerm/reproseries documentation built on Jan. 9, 2020, 1:27 a.m.