opennaijR simplifies access to Nigerian open data from the CBN, NBS, NGX, and other sources. It automates the cleaning of official datasets, turning hours of manual work into seconds of analysis-ready results. It is a reproducible macroeconomic data engineering for Nigeria. Use discover_datasets() to find your variable of interest, then pass it to cbn() to start downloading.
Installation
#Install released version from GitHub using remotes:
install.packages("remotes")
remotes::install_github("laws2020/opennaijR")Usage
Search for available datasets
Before pulling data with cbn(), you need to know what is available. Use discover_datasets(), it is the entry point to opennaijR. Users should never guess dataset names.
It returns:
- available datasets exist
- their source (CBN, WB, NBS, etc.)
- available indicators
discover_datasets()|> print(n = 6)## # A tibble: 23 × 5
## dataset_key dataset_id source aliases variables
## <chr> <chr> <chr> <chr> <chr>
## 1 crude_oil cbn_crude_oil CBN crude oil, oil price, bo… price_bo…
## 2 daily_crude_oil cbn_daily_crude_oil CBN daily crude oil, daily o… price_bo…
## 3 discount_rates cbn_discount_rates CBN discount rates, discount… discount…
## 4 ntb_cbn cbn_ntb CBN ntb, nigeria treasury bi… total_su…
## 5 fgn_bond cbn_fgn_bond CBN fgn bond, fgn bonds, nig… total_su…
## 6 omo cbn_omo CBN omo, open market operati… total_su…
## # ℹ 17 more rowsAlternative: print everything:
discover_datasets() |> print(n = Inf)Download the Data
Once you have found a dataset name using discover_datasets(), pass it to cbn() to download the data into your R session. You can use either the dataset_key or an alias. For example, to retrieve official inflation data from the Central Bank of Nigeria using the “inflation” alias:
# Fetch data (cached internally by opennaijR after the first run)
infl <- cbn("inflation")View the records
## <opennaijR table>
## Rows: 6 Columns: 4
##
## date headline_yoy food_yoy core_ex_farm_yoy
## 1 2026-01-01 15.10 8.89 17.18
## 2 2025-12-01 15.15 10.84 18.16
## 3 2025-11-01 17.33 14.21 19.84
## 4 2025-10-01 18.97 16.30 20.61
## 5 2025-09-01 20.98 20.16 21.61
## 6 2025-08-01 23.14 25.30 22.63That’s it. infl is now a clean, tidy data frame ready for analysis—no manual cleaning required.
Under the hood, cbn() connects to the official CBN data stream, normalises column names, parses dates into proper R Date objects, and returns a structure compatible with ggplot2 and dplyr.
infl is ready for lm(), ggplot2, dplyr, or any other tool you reach for. The next sections show the opennaijR-specific functions that make your workflow more precise and reproducible.
Cache Management
Data is fetched from the web once, then stored locally for near-instant loading in future sessions.
# First run: Downloads from CBN | Subequent runs: Loads from local disk
cbn("inflation")[, c("date", "headline_yoy", "food_yoy", "core_ex_farm_yoy")]## 📦 Loading cached CBN data from pins
## <opennaijR table>
## Rows: 277 Columns: 4
##
## date headline_yoy food_yoy core_ex_farm_yoy
## 1 2026-01-01 15.10 8.89 17.18
## 2 2025-12-01 15.15 10.84 18.16
## 3 2025-11-01 17.33 14.21 19.84
## 4 2025-10-01 18.97 16.30 20.61
## 5 2025-09-01 20.98 20.16 21.61
## 6 2025-08-01 23.14 25.30 22.63
## 7 2025-07-01 24.94 26.20 23.93
## 8 2025-06-01 26.06 24.55 25.88
## 9 2025-05-01 26.06 24.55 25.88
## 10 2025-04-01 26.82 24.68 27.12Take Control
Manage your local cache with the naijr_cache_* family:
#Use code with caution.
naijr_cache_clear("inflation") #Force a fresh download by wiping specific datasets.
naijr_cache_list() #See every dataset currently stored on your machine.
naijr_cache_info() #Check cache size and file locations.Schema Control — apply_projection()
cbn() returns all available columns. apply_projection() lets you select, rename, and reorder them in a single auditable call.
infl <- cbn("inflation")
proj <- apply_projection(
infl,
cols = c("date", "headline_yoy", "food_yoy", "core_ex_farm_yoy"),
rename = c(
headline = "headline_yoy",
food = "food_yoy",
core = "core_ex_farm_yoy"
),
order = c("date", "headline", "food", "core"),
reason = "Key inflation measures with clean names"
)
head(proj)## <opennaijR table>
## Rows: 6 Columns: 4
##
## date headline food core
## 1 2026-01-01 15.10 8.89 17.18
## 2 2025-12-01 15.15 10.84 18.16
## 3 2025-11-01 17.33 14.21 19.84
## 4 2025-10-01 18.97 16.30 20.61
## 5 2025-09-01 20.98 20.16 21.61
## 6 2025-08-01 23.14 25.30 22.63Every call leaves a projection manifest — a built-in record of what changed and why:
attr(proj, "projection_manifest")## [[1]]
## [[1]]$timestamp
## [1] "2026-03-02 05:06:38 WAT"
##
## [[1]]$action
## [1] "apply_projection"
##
## [[1]]$filter
## NULL
##
## [[1]]$kept
## [1] "date" "headline_yoy" "food_yoy" "core_ex_farm_yoy"
##
## [[1]]$renamed
## headline food core
## "headline_yoy" "food_yoy" "core_ex_farm_yoy"
##
## [[1]]$ordered
## [1] "date" "headline" "food" "core"
##
## [[1]]$reason
## [1] "Key inflation measures with clean names"The manifest records the timestamp, which columns were kept, how they were renamed, and the reason you supplied. This makes your workflow reproducible by design.
Feature Engineering — derive_measure()
Create new analytical columns directly from your dataset. Reference column names as bare expressions — no $ or [[]] needed.
infl_features <- derive_measure(
infl,
gap = headline_yoy - food_yoy,
accelerating = headline_yoy > lag(headline_yoy),
high_regime = headline_yoy > 15,
reason = "Inflation diagnostic indicators"
)
head(infl_features)[, c("date", "gap", "accelerating","high_regime" )]## <opennaijR table>
## Rows: 6 Columns: 4
##
## date gap accelerating high_regime
## 1 2026-01-01 6.21 FALSE TRUE
## 2 2025-12-01 4.31 FALSE TRUE
## 3 2025-11-01 3.12 FALSE TRUE
## 4 2025-10-01 2.67 FALSE TRUE
## 5 2025-09-01 0.82 FALSE TRUE
## 6 2025-08-01 -2.16 FALSE TRUELike projections, every derivation is tracked:
attr(infl_features, "derive_manifest")## [[1]]
## [[1]]$timestamp
## [1] "2026-03-02 05:06:38 WAT"
##
## [[1]]$action
## [1] "derive_measure"
##
## [[1]]$derived
## [1] "gap" "accelerating" "high_regime"
##
## [[1]]$expressions
## gap accelerating
## "~headline_yoy - food_yoy" "~headline_yoy > lag(headline_yoy)"
## high_regime
## "~headline_yoy > 15"
##
## [[1]]$reason
## [1] "Inflation diagnostic indicators"The manifest stores the exact expressions evaluated, the timestamp, and your declared intent. You can chain multiple derive_measure() calls and the record accumulates step by step — giving you a full audit trail from raw data to final feature set.
A Complete Workflow
library(opennaijR)
# 1. Find and fetch
infl <- cbn("inflation")
# 2. Clean schema
infl_clean <- apply_projection(
infl,
cols = c("date", "headline_yoy", "food_yoy", "core_ex_farm_yoy"),
rename = c(Date = "date", Headline = "headline_yoy",
Food = "food_yoy", Core = "core_ex_farm_yoy"),
reason = "Baseline schema for analysis"
)
# 3. Engineer features
infl_ready <- derive_measure(
infl_clean,
gap = Headline - Food,
accelerating = Headline > lag(Headline),
high_regime = Headline > 15,
reason = "Diagnostic indicators"
)
# 4. Model
lm(Headline ~ Food + Core, data = infl_ready) |> summary()Learn More
Full documentation and worked examples are available in the package articles:
-
Schema Management — a complete guide to
apply_projection() -
Feature Engineering — every class of derivation with
derive_measure() - Macro Regime & Regression Analysis — a full research pipeline from raw data to econometric models