Source: /home/coolhand/html/datavis/data_trove/data/linguistic/wals_parameters.csv
Saturn profiled 192 rows across 5 columns. The stats below are deterministic and machine-readable; the prose is a language-model interpretation of those stats (opt-in, added after the fact, never sees raw rows).
!pip install saturn-dissect
import subprocess
subprocess.run([
"saturn", "analyze", "/home/coolhand/html/datavis/data_trove/data/linguistic/wals_parameters.csv",
"--findings", "data-trove-wals-world-atlas-of-language-structures.json",
"--llm", "anthropic:default",
])This dataset is a catalogue of 192 linguistic parameters from the World Atlas of Language Structures (WALS), each representing a distinct typological feature of human languages (e.g., 'Consonant Inventories', 'Vowel Quality Inventories'). Every row is uniquely identified by both an ID (e.g., '1A', '2A') and a Name, meaning there are no duplicates or groupings to aggregate within those columns. The most analytically interesting column is Chapter_ID, which groups these 192 parameters into 144 chapters — indicating that some chapters contain multiple parameters worth investigating. The Chapter_ID distribution is fairly uniform (mean ~84.5, median ~89.5, near-symmetric with slight left skew), suggesting chapters are spread across the full range with no heavy clustering.
citing: row_count · column_count · n_unique · null_rate · mean · median · skew · max · min · n_unique
saturn.schema() · 5 columns
| column | kind | n | null% | unique | alerts |
|---|---|---|---|---|---|
| ID | categorical | 192 | 0.0% | 192 | long_tail |
| Name | categorical | 192 | 0.0% | 192 | long_tail |
| Description | unknown | 192 | 0.0% | — | skipped |
| ColumnSpec | unknown | 192 | 0.0% | — | skipped |
| Chapter_ID | numeric | 192 | 0.0% | 144 |
| bin | count |
|---|---|
| 1 – 12 | 12 |
| 12 – 23 | 12 |
| 23 – 34 | 12 |
| 34 – 45 | 12 |
| 45 – 56 | 11 |
| 56 – 67 | 12 |
| 67 – 78 | 11 |
| 78 – 89 | 13 |
| 89 – 100 | 17 |
| 100 – 111 | 13 |
| 111 – 122 | 11 |
| 122 – 133 | 12 |
| 133 – 144 | 44 |
| value | count | share |
|---|---|---|
| Consonant Inventories | 1 | 0.5% |
| Vowel Quality Inventories | 1 | 0.5% |
| Consonant-Vowel Ratio | 1 | 0.5% |
| Voicing in Plosives and Fricatives | 1 | 0.5% |
| Voicing and Gaps in Plosive Systems | 1 | 0.5% |
| Uvular Consonants | 1 | 0.5% |
| Glottalized Consonants | 1 | 0.5% |
| Lateral Consonants | 1 | 0.5% |
| The Velar Nasal | 1 | 0.5% |
| Vowel Nasalization | 1 | 0.5% |
| Nasal Vowels in West Africa | 1 | 0.5% |
| Front Rounded Vowels | 1 | 0.5% |
| Syllable Structure | 1 | 0.5% |
| Tone | 1 | 0.5% |
| Fixed Stress Locations | 1 | 0.5% |
| Weight-Sensitive Stress | 1 | 0.5% |
| Weight Factors in Weight-Sensitive Stress Systems | 1 | 0.5% |
| Rhythm Types | 1 | 0.5% |
| Absence of Common Consonants | 1 | 0.5% |
| Presence of Uncommon Consonants | 1 | 0.5% |
| value | count | share |
|---|---|---|
| 1A | 1 | 0.5% |
| 2A | 1 | 0.5% |
| 3A | 1 | 0.5% |
| 4A | 1 | 0.5% |
| 5A | 1 | 0.5% |
| 6A | 1 | 0.5% |
| 7A | 1 | 0.5% |
| 8A | 1 | 0.5% |
| 9A | 1 | 0.5% |
| 10A | 1 | 0.5% |
| 10B | 1 | 0.5% |
| 11A | 1 | 0.5% |
| 12A | 1 | 0.5% |
| 13A | 1 | 0.5% |
| 14A | 1 | 0.5% |
| 15A | 1 | 0.5% |
| 16A | 1 | 0.5% |
| 17A | 1 | 0.5% |
| 18A | 1 | 0.5% |
| 19A | 1 | 0.5% |
| column | kind | null % |
|---|---|---|
| ID | categorical | 0.0% |
| Name | categorical | 0.0% |
| Description | unknown | 0.0% |
| ColumnSpec | unknown | 0.0% |
| Chapter_ID | numeric | 0.0% |
This column is a row identifier, with 192 unique values across 192 rows, zero nulls, and a maximum entropy ratio of 1.0 — every value appears exactly once. The alphanumeric pattern (e.g., '1A', '2A', …, '10A') suggests a structured code system, possibly denoting sequential items with a suffix category. The long_tail alert is a statistical artifact of perfect uniqueness rather than a genuine distributional concern.
Treatment: Exclude from modelling features; retain as a row key for traceability or join operations.
saturn.columns["ID"].stats
| stat | value |
|---|---|
| n | 192 |
| nulls | 0 (0.0%) |
| unique | 192 |
| top_value | 1A |
| top_rate | 0.005208 |
| cardinality | 192 |
| entropy | 7.585 |
| entropy_ratio | 1 |
| alert: long_tail | 192 singleton categories |
ID.
| value | count | share |
|---|---|---|
| 1A | 1 | 0.5% |
| 2A | 1 | 0.5% |
| 3A | 1 | 0.5% |
| 4A | 1 | 0.5% |
| 5A | 1 | 0.5% |
| 6A | 1 | 0.5% |
| 7A | 1 | 0.5% |
| 8A | 1 | 0.5% |
| 9A | 1 | 0.5% |
| 10A | 1 | 0.5% |
| 10B | 1 | 0.5% |
| 11A | 1 | 0.5% |
| 12A | 1 | 0.5% |
| 13A | 1 | 0.5% |
| 14A | 1 | 0.5% |
| 15A | 1 | 0.5% |
| 16A | 1 | 0.5% |
| 17A | 1 | 0.5% |
| 18A | 1 | 0.5% |
| 19A | 1 | 0.5% |
This column contains the names of linguistic typology features or chapters, almost certainly from the World Atlas of Language Structures (WALS) or a similar comparative linguistics dataset. Every one of the 192 rows has a unique name (cardinality 192, entropy_ratio 1.0), meaning this column is a perfect natural-language identifier with no repetition. The top values — 'Consonant Inventories', 'Vowel Quality Inventories', 'Consonant-Vowel Ratio' — confirm these are named phonological/typological feature categories, not free-form text.
Treatment: Use as a human-readable row label or index; do not encode as a categorical feature — treat as an identifier or join key on feature name.
saturn.columns["Name"].stats
| stat | value |
|---|---|
| n | 192 |
| nulls | 0 (0.0%) |
| unique | 192 |
| top_value | Consonant Inventories |
| top_rate | 0.005208 |
| cardinality | 192 |
| entropy | 7.585 |
| entropy_ratio | 1 |
| alert: long_tail | 192 singleton categories |
Name.
| value | count | share |
|---|---|---|
| Consonant Inventories | 1 | 0.5% |
| Vowel Quality Inventories | 1 | 0.5% |
| Consonant-Vowel Ratio | 1 | 0.5% |
| Voicing in Plosives and Fricatives | 1 | 0.5% |
| Voicing and Gaps in Plosive Systems | 1 | 0.5% |
| Uvular Consonants | 1 | 0.5% |
| Glottalized Consonants | 1 | 0.5% |
| Lateral Consonants | 1 | 0.5% |
| The Velar Nasal | 1 | 0.5% |
| Vowel Nasalization | 1 | 0.5% |
| Nasal Vowels in West Africa | 1 | 0.5% |
| Front Rounded Vowels | 1 | 0.5% |
| Syllable Structure | 1 | 0.5% |
| Tone | 1 | 0.5% |
| Fixed Stress Locations | 1 | 0.5% |
| Weight-Sensitive Stress | 1 | 0.5% |
| Weight Factors in Weight-Sensitive Stress Systems | 1 | 0.5% |
| Rhythm Types | 1 | 0.5% |
| Absence of Common Consonants | 1 | 0.5% |
| Presence of Uncommon Consonants | 1 | 0.5% |
This column contains textual descriptions with 192 non-null rows and a null rate of 0.0%, suggesting complete population coverage. Profiling was skipped entirely (alert: 'skipped'), so no uniqueness, frequency, or length statistics are available — the column's content distribution is opaque. The 'unknown' kind designation indicates the profiler could not classify it, which is common for free-text or mixed-content fields. Downstream treatment should proceed cautiously given the absence of any statistical evidence.
Treatment: Inspect raw values manually, then tokenize and embed or apply NLP preprocessing before modelling.
saturn.columns["Description"].stats
| stat | value |
|---|---|
| n | 192 |
| nulls | 0 (0.0%) |
| unique | — |
| alert: skipped | no profiler for kind=unknown |
The column 'ColumnSpec' was skipped by the profiler, yielding no distributional statistics, cardinality, or type classification beyond 192 non-null rows. Without further evidence it is impossible to characterise the column's content, range, or role. The complete absence of stats alongside the 'skipped' alert is the surprising signal here — likely due to an unsupported or complex data type (e.g. nested struct, binary, or array).
Treatment: Inspect raw values manually to determine type and content before assigning a profiling strategy or modelling treatment.
saturn.columns["ColumnSpec"].stats
| stat | value |
|---|---|
| n | 192 |
| nulls | 0 (0.0%) |
| unique | — |
| alert: skipped | no profiler for kind=unknown |
This column is a chapter identifier, likely a foreign or primary key linking rows to one of 144 distinct chapters across a book or structured document. With 192 rows but only 144 unique values, each chapter ID appears on average 1.33 times, indicating some chapters have multiple associated records. The near-uniform distribution (kurtosis –1.26, skew –0.19, IQR spanning almost the full 1–144 range) suggests IDs are assigned sequentially and coverage across chapters is fairly even, though the duplication pattern warrants checking whether it reflects intentional one-to-many relationships or data quality issues.
Treatment: Left-join on this ID to the chapters reference table; investigate the 48 duplicate entries to confirm they represent legitimate one-to-many relationships.
saturn.columns["Chapter_ID"].stats
| stat | value |
|---|---|
| n | 192 |
| nulls | 0 (0.0%) |
| unique | 144 |
| min | 1 |
| max | 144 |
| mean | 84.52 |
| median | 89.5 |
| std | 45.75 |
| q1 | 44.75 |
| q3 | 129.2 |
| iqr | 84.5 |
| skew | -0.195 |
| kurtosis | -1.262 |
| n_outliers | 0 |
| outlier_rate | 0 |
| zero_rate | 0 |
Chapter_ID. Vertical dash marks the median.
| bin | count |
|---|---|
| 1 – 12 | 12 |
| 12 – 23 | 12 |
| 23 – 34 | 12 |
| 34 – 45 | 12 |
| 45 – 56 | 11 |
| 56 – 67 | 12 |
| 67 – 78 | 11 |
| 78 – 89 | 13 |
| 89 – 100 | 17 |
| 100 – 111 | 13 |
| 111 – 122 | 11 |
| 122 – 133 | 12 |
| 133 – 144 | 44 |
click to copy
@misc{saturn-data-trove-wals-world-atlas-of-language-structures-2026,
author = {Steuber, Luke},
title = {Saturn reading: data trove wals world atlas of language structures},
year ={2026},
howpublished = {\url{https://dr.eamer.dev/saturn/view/data-trove-wals-world-atlas-of-language-structures}},
note = {Profiled with saturn-dissect v0.2.0, prompt saturn-insight-v2, model anthropic:default},
}
Steuber, L. (2026). Saturn reading: data trove wals world atlas of language structures. Source: /home/coolhand/html/datavis/data_trove/data/linguistic/wals_parameters.csv. Profiled with saturn-dissect v0.2.0 (saturn-insight-v2, anthropic:default). Retrieved from https://dr.eamer.dev/saturn/view/data-trove-wals-world-atlas-of-language-structures