electricsheepafrica/africa-who-number-of-neonatal-deaths

Africa — WHO GHO: Number of neonatal deaths

Indicator code: CM_03 HuggingFace slug: electricsheepafrica/africa-who-number-of-neonatal-deaths Source: WHO Global Health Observatory License: CC BY 4.0 — WHO Open Data

---

Dataset Description

This dataset contains country-level observations for the WHO GHO indicator "Number of neonatal deaths" (CM_03) across African nations, spanning 1955–2023. It is part of the Electric Sheep Africa collection — a unified, ML-ready repository of African data.

Data is sourced directly from the WHO Global Health Observatory OData API and repackaged as Parquet files with a consistent schema. All values are drawn from NumericValue (the float-precision field), not the display string. Confidence interval bounds (value_low, value_high) are included where available.

---

Coverage

Countries	47 African nations
Years	1955 – 2023
Total rows	2,414
Region filter	WHO AFRO (ParentLocationCode = 'AFR')

Countries included: AGO, BDI, BEN, BFA, BWA, CAF, CIV, CMR, COD, COG, COM, CPV, DZA, ERI, ETH, GAB, GHA, GIN, GMB, GNB … and 27 more

---

Sub-dimensions

• SEX: SEX_BTSX

When an indicator is stratified (e.g., by sex or age group), each unique combination of country × year × dimension produces a separate row. Filter on dim1 / dim2 for the stratum you need, or aggregate across strata.

---

Schema

Column	Type	Description
indicator_code	string	GHO indicator code (e.g., CM_03)
country_iso3	string	ISO 3166-1 alpha-3 country code
who_region	string	WHO region code (always AFR here)
year	int	Observation year
value_numeric	float	Point estimate (primary ML target)
value_low	float	Lower confidence bound (if available)
value_high	float	Upper confidence bound (if available)
value_display	string	Formatted display string, e.g. "58.3 [57.7–59.0]"
dim1_type	string	Dimension 1 type, e.g. SEX, RESIDENCEAREATYPE
dim1	string	Dimension 1 value, e.g. SEX_BTSX, RURAL
dim2_type	string	Dimension 2 type (if present)
dim2	string	Dimension 2 value (if present)
last_updated	string	WHO data last-updated timestamp

---

Usage

from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-who-number-of-neonatal-deaths")
df = ds["train"].to_pandas()

# Both-sexes, national level only
national = df[df.get("dim1", "").str.endswith("_BTSX") | df.get("dim1", pd.Series()).isna()]

# Time series for one country
kenya = df[df["country_iso3"] == "KEN"].sort_values("year")

---

Citation

@misc{who_gho_cm_03,
  title     = {WHO Global Health Observatory: Number of neonatal deaths},
  author    = {World Health Organization},
  year      = {2023},
  url       = {https://www.who.int/data/gho/data/indicators/indicator-details/GHO/CM_03},
  note      = {Repackaged by Electric Sheep Africa}
}

---

Repackaged by Electric Sheep Africa from WHO GHO open data. Original data © World Health Organization, licensed CC BY 4.0.