README.md

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: other
multilinguality:
- monolingual
size_categories:
- 1K<n<10K
source_datasets:
- original
task_categories:
- tabular-classification
- tabular-regression
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- hxl
- indicators
- tkm
pretty_name: "Turkmenistan - Historical Health Indicators"
dataset_info:
  splits:
    - name: train
      num_examples: 7921
    - name: test
      num_examples: 1980
---

# Turkmenistan - Historical Health Indicators

**Publisher:** World Health Organization · **Source:** [HDX](https://data.humdata.org/dataset/who-historical-data-for-tkm) · **License:** `hdx-other` · **Updated:** 2025-02-07

---

## Abstract

This dataset contains historical data from WHO's [data portal](https://www.who.int/gho/en/).

Each row in this dataset represents first-level administrative unit observations. Data was last updated on HDX on 2025-02-07. Geographic scope: **TKM**.

*Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).*

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Humanitarian and development data |
| **Unit of observation** | First-level administrative unit observations |
| **Rows (total)** | 9,902 |
| **Columns** | 18 (6 numeric, 12 categorical, 0 datetime) |
| **Train split** | 7,921 rows |
| **Test split** | 1,980 rows |
| **Geographic scope** | TKM |
| **Publisher** | World Health Organization |
| **HDX last updated** | 2025-02-07 |

---

## Variables

**Geographic** — `gho_display` (Mean BMI (kg/m&#xb2;) (crude estimate), Adolescent mortality rate (per 1 000 age specific cohort), Alcohol, recorded per capita (15+) consumption (in litres of pure alcohol)), `year_display` (range 1970.0–2025.0), `startyear` (range 1970.0–2025.0), `endyear` (range 1970.0–2025.0), `region_code` (EUR, #region+code) and 4 others.

**Outcome / Measurement** — `value` (No data, No, Yes).

**Identifier / Metadata** — `gho_code` (NCD_BMI_MEANC, CHILDMORT10TO19, SA_0000001400_ARCHIVED), `dimension_code` (SEX_FMLE, SEX_MLE, SEX_BTSX), `dimension_name` (Female, Male, Both sexes), `esa_source`, `esa_processed`.

**Other** — `numeric` (range 0.0–266862800.0), `low` (range 0.0–37051.0), `high` (range 0.0–173952.0).

---

## Quick Start

```python
from datasets import load_dataset

ds    = load_dataset("electricsheepafrica/asia-who-historical-data-for-turkmenistan")
train = ds["train"].to_pandas()
test  = ds["test"].to_pandas()

print(train.shape)
train.head()
```

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `gho_code` | object | 0.0% | NCD_BMI_MEANC, CHILDMORT10TO19, SA_0000001400_ARCHIVED |
| `gho_display` | object | 0.0% | Mean BMI (kg/m&#xb2;) (crude estimate), Adolescent mortality rate (per 1 000 age specific cohort), Alcohol, recorded per capita (15+) consumption (in litres of pure alcohol) |
| `year_display` | float64 | 0.0% | 1970.0 – 2025.0 (mean 2009.9776) |
| `startyear` | float64 | 0.0% | 1970.0 – 2025.0 (mean 2009.9657) |
| `endyear` | float64 | 0.0% | 1970.0 – 2025.0 (mean 2009.9776) |
| `region_code` | object | 0.0% | EUR, #region+code |
| `region_display` | object | 0.0% | Europe, #region+name |
| `country_code` | object | 0.0% | TKM, #country+code |
| `country_display` | object | 0.0% | Turkmenistan, #country+name |
| `dimension_type` | object | 19.2% | SEX, WEALTHDECILE, WEALTHQUINTILE |
| `dimension_code` | object | 19.2% | SEX_FMLE, SEX_MLE, SEX_BTSX |
| `dimension_name` | object | 20.4% | Female, Male, Both sexes |
| `numeric` | float64 | 34.8% | 0.0 – 266862800.0 (mean 41711.3983) |
| `value` | object | 1.3% | No data, No, Yes |
| `low` | float64 | 49.8% | 0.0 – 37051.0 (mean 114.7989) |
| `high` | float64 | 49.9% | 0.0 – 173952.0 (mean 289.3685) |
| `esa_source` | object | 0.0% |  |
| `esa_processed` | object | 0.0% |  |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `year_display` | 1970.0 | 2025.0 | 2009.9776 | 2013.0 |
| `startyear` | 1970.0 | 2025.0 | 2009.9657 | 2012.0 |
| `endyear` | 1970.0 | 2025.0 | 2009.9776 | 2013.0 |
| `numeric` | 0.0 | 266862800.0 | 41711.3983 | 34.5564 |
| `low` | 0.0 | 37051.0 | 114.7989 | 25.7971 |
| `high` | 0.0 | 173952.0 | 289.3685 | 44.5 |

---

## Curation

Raw data was downloaded from HDX via the CKAN API and converted to Parquet. Column names were lowercased and standardised to snake_case. Common missing-value markers (`N/A`, `null`, `none`, `-`, `unknown`, `no data`, `#N/A`) were unified to `NaN`. 1 column(s) with >80% missing values were removed: `gho_url`. 67 exact duplicate rows were removed. 6 column(s) were cast from string to numeric or datetime based on parse-success rate (>85% threshold). The dataset was split 80/20 into train and test partitions using a fixed random seed (42) and saved as Snappy-compressed Parquet.

---

## Limitations

- Data originates from World Health Organization and has not been independently validated by ESA.
- Automated cleaning cannot correct for misreported values, definitional inconsistencies, or sampling bias in the original collection.
- The following columns have >20% missing values and should be treated with caution in modelling: `dimension_name`, `numeric`, `low`, `high`.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/who-historical-data-for-tkm) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_asia_who_historical_data_for_turkmenistan,
  title     = {Turkmenistan - Historical Health Indicators},
  author    = {World Health Organization},
  year      = {2025},
  url       = {https://data.humdata.org/dataset/who-historical-data-for-tkm},
  note      = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}
```

---

*[Electric Sheep Africa](https://huggingface.co/electricsheepafrica) — Africa's ML dataset infrastructure. Lagos, Nigeria.*