Skip to content

Latest commit

 

History

History
57 lines (34 loc) · 3.42 KB

readme.md

File metadata and controls

57 lines (34 loc) · 3.42 KB
Raw

Macro-éCHo

Matthias Spichiger - [email protected] Based on Daniel Kaufmann, émissions "Macro-éCHo", Université de Neuchâtel [email protected] readme.md for replication files

Changes

  • Following Bryan, Jennifer, et al., I use an R project to make the file paths shorter and independent from the user .

  • Following Wickham, Hadley, I use the tidyverse to import the data , hence base::read.csv() became readr::read_delim() and xlsx::read.xlsx() became readxl::read_excel().

    Note that the tidyverse functions load data as tibble instead of data.frame objects. When a single column of a tibble is selected, it remains a column-vector and is not transposed to a row-vector, meaning you have to use dplyr::pull() before applying functions that use row-vectors as input.

  • The Core tidyverse packages (dplyr, forcats, ggplot2, lubridate, purrr, readr, stringr, tibble, and tidyr) are loaded in one go with library(tidyverse). Note that the non-core tidyverse package readxl has to be loaded separately with load(readxl) Packages that are not used in the scripts are not loaded.

  • If the links to the data have changed they have been updated.

  • tsbox::ts_ggplot() is not working and has been replaced with ggplot2::ggplot().

To create data frames for ggplot2::ggplot(), time series objects are collected within tsbox::ts_c() and converted to a data frame in long format with tsbox::ts_df().

  • If the time stamps of two series do not match, the series are converted from xts to ts objects in order to make implicit missing values explicit as NAs. Before plotting, those NAs are deleted with base::na.omit().

  • In all plots, I include the years on the x-axes, and I add a dashed horizontal line at zero.

  • The data source and the creator (@econmaett) are included in the caption.

  • Inside of ggplot functions, the deprecated size-argument is replaced with linewidth (See Pedersen, Thomas Lin).

  • Legends are included in the title, using the ggtext package.

  • I replaced ---- with **** so that the dashes indicate the document outline (Ctrl + Shift + O).

  • I use the styler package to format the scripts according to the tidyverse style guide.

Notes

These files replicate the charts used in my video series "Macro-éCHo". All data sources can be inferred from the codes. All files can be freely downloaded and distributed. Use them at your own risk with no guarantee or maintenance. If you spot an error, thanks for letting me know.

The videos are available on www.youtube.com/channel/UCJpACBsnn1eQTObWz5LniGg

Episodes Season 1

  • S01E01: Prime de risque: "La courbe de fièvre de l'économie Suisse"
  • S01E02: Baromètre conjoncturel mondial: "Les saints de glace jettent un froid sur la Suisse"
  • S01E03: Le taux dirécteur: "Les taux d'intérêts: Est-ce qu'ils les instructions du dirécteur?"
  • S01E04: Les prix: "Les prix baissent en Suisse – C’est top n’est-ce pas?"
  • S01E05: Prime de risque Pt2: "La courbe de fièvre de l'économie Suisse - Partie 2" Note that the files for this episode are available on github.com/dankaufmann/f-curve