---
title: "Get Started with tidyBdE"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Get Started with tidyBdE}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

<!-- tidyBdE.Rmd is generated from tidyBdE.Rmd.orig. Please edit that file -->



**tidyBdE** is an API package that helps to retrieve data from [Banco de
España](https://www.bde.es/webbe/en/estadisticas/recursos/descargas-completas.html).
The data is returned as a [tibble](https://tibble.tidyverse.org/), and the
package automatically infers the format of each time-series (dates, characters,
and numbers).

## Search series

Banco de España (**BdE**) provides several time-series, either produced by the
institution itself or compiled from other sources, such as
[Eurostat](https://ec.europa.eu/eurostat) or [INE](https://www.ine.es/).

The basic entry points for searching time-series are the catalogs (*indexes*).
You can search for any series by name:


``` r
library(tidyBdE)

library(ggplot2)
library(dplyr)
library(tidyr)


# Search GBP on "TC" (exchange rate) catalog
xr_gbp <- bde_catalog_search("GBP", catalog = "TC")

xr_gbp |>
  select(Numero_secuencial, Descripcion_de_la_serie) |>
  # To table on document
  knitr::kable()
```



| Numero_secuencial|Descripcion_de_la_serie                                            |
|-----------------:|:------------------------------------------------------------------|
|            573214|Tipo de cambio. Libras esterlinas por euro (GBP/EUR).Datos diarios |



**Note:** BdE files are currently provided only in Spanish, as the institution
works on an English version. Search terms should be provided in Spanish to
obtain results.

Once you have found your series, you can load the GBP/EUR exchange rate using
the sequential number reference (`Numero_Secuencial`):


``` r
seq_number <- xr_gbp |>
  # First record
  slice(1) |>
  # Get the ID
  select(Numero_secuencial) |>
  # Convert to numeric
  as.double()


seq_number
#> [1] 573214


time_series <- bde_series_load(seq_number, series_label = "EUR_GBP_XR") |>
  filter(Date >= "2010-01-01" & Date <= "2020-12-31") |>
  drop_na()
```

## Plot series

The package also provides a custom **ggplot2** theme based on BdE's
publications:


``` r
ggplot(time_series, aes(x = Date, y = EUR_GBP_XR)) +
  geom_line(colour = bde_tidy_palettes(n = 1)) +
  geom_smooth(method = "gam", colour = bde_tidy_palettes(n = 2)[2]) +
  labs(
    title = "EUR/GBP Exchange Rate (2010-2020)",
    subtitle = "%",
    caption = "Source: BdE"
  ) +
  geom_vline(
    xintercept = as.Date("2016-06-23"),
    linetype = "dotted"
  ) +
  geom_label(aes(
    x = as.Date("2016-06-23"),
    y = 0.95,
    label = "Brexit"
  )) +
  coord_cartesian(ylim = c(0.7, 1)) +
  theme_tidybde()
```

<div class="figure">
<img src="./chart-1.png" alt="EUR/GBP Exchange Rate (2010-2020)" width="100%" />
<p class="caption">EUR/GBP Exchange Rate (2010-2020)</p>
</div>

The package also provides convenience functions for a selection of the most
relevant macroeconomic series, eliminating the need for manual searching:


``` r
# Data in "long" format

plotseries <- bde_ind_gdp_var("GDP YoY", out_format = "long") |>
  bind_rows(
    bde_ind_unemployment_rate("Unemployment Rate", out_format = "long")
  ) |>
  drop_na() |>
  filter(Date >= "2010-01-01" & Date <= "2019-12-31")

ggplot(plotseries, aes(x = Date, y = serie_value)) +
  geom_line(aes(color = serie_name), linewidth = 1) +
  labs(
    title = "Spanish Economic Indicators (2010-2019)",
    subtitle = "%",
    caption = "Source: BdE"
  ) +
  theme_tidybde() +
  scale_color_bde_d(palette = "bde_vivid_pal") # Custom palette on the package
```

<div class="figure">
<img src="./macroseries-1.png" alt="Spanish Economic Indicators (2010-2019)" width="100%" />
<p class="caption">Spanish Economic Indicators (2010-2019)</p>
</div>

## A note on caching

You can use **tidyBdE** to create your own local repository at a given local
directory passing the following option:


``` r
options(bde_cache_dir = "./path/to/location")
```

When this option is set, **tidyBdE** will look for cached files in the
`bde_cache_dir` directory and load them, speeding up data retrieval.

It is possible to update the data (i.e. after every monthly or quarterly data
release) with the following commands:


``` r
bde_catalog_update()

# Or use update_cache = TRUE in most functions

bde_series_load("SOME ID", update_cache = TRUE)
```