processed cognate sets

saturn notebook · generated 2026-05-01 Report Notebook

Overview

Source: /home/coolhand/servers/diachronica/etymology_atlas/processed/cognate_sets.json

Saturn profiled 4,981 rows across 8 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).

[2]:

!pip install saturn-dissect
import subprocess
subprocess.run([
    "saturn", "analyze", "/home/coolhand/servers/diachronica/etymology_atlas/processed/cognate_sets.json",
    "--findings", "processed-cognate_sets.json",
    "--llm", "anthropic:claude-opus-4-7",
])

Summary confidence: high

This dataset contains 4,981 cognate sets sourced entirely from the 'iecor' source_dataset, each identified by a unique cognate_id. The two main numeric signals are language_count and word_count, which are nearly identical in distribution: both have a median of 2 and mean around 5.17, but stretch out to a maximum of 157 with skew above 6.8 and roughly 13% of rows flagged as outliers. That long tail is the most interesting story — most cognate sets are small, but a minority span very many languages/words and deserve a closer look. Note that concept is empty for every row, confidence is constant at 1.0, and source_dataset has only one value, so those columns carry no analytic signal.

citing: row_count · column_count · language_count · word_count · concept · confidence · source_dataset · cognate_id

Out[4]:

saturn.schema() · 8 columns

column	kind	n	null%	unique	alerts
cognate_id	text	4,981	0.0%	4,981	near_unique one_word short_text
concept	categorical	4,981	0.0%	1	imbalance
words	unknown	4,981	0.0%	—	skipped
source_dataset	categorical	4,981	0.0%	1	imbalance
confidence	numeric	4,981	0.0%	1	constant
word_count	numeric	4,981	0.0%	93	high_skew outliers
language_count	numeric	4,981	0.0%	94	high_skew outliers
sources	unknown	4,981	0.0%	—	skipped

Fig 1.

language_count · Look at the long right tail — most sets cover just 1-2 languages but a few reach 157.

Show data table

Histogram bins for language_count (median: 2.0).
bin	count
1 – 4.9	3849
4.9 – 8.8	483
8.8 – 12.7	194
12.7 – 16.6	96
16.6 – 20.5	90
20.5 – 24.4	107
24.4 – 28.3	32
28.3 – 32.2	25
32.2 – 36.1	15
36.1 – 40	4
40 – 43.9	8
43.9 – 47.8	8
47.8 – 51.7	6
51.7 – 55.6	5
55.6 – 59.5	3
59.5 – 63.4	3
63.4 – 67.3	8
67.3 – 71.2	3
71.2 – 75.1	3
75.1 – 79	4
79 – 82.9	0
82.9 – 86.8	1
86.8 – 90.7	2
90.7 – 94.6	4
94.6 – 98.5	5
98.5 – 102.4	2
102.4 – 106.3	3
106.3 – 110.2	2
110.2 – 114.1	2
114.1 – 118	3
118 – 121.9	0
121.9 – 125.8	1
125.8 – 129.7	1
129.7 – 133.6	0
133.6 – 137.5	2
137.5 – 141.4	2
141.4 – 145.3	0
145.3 – 149.2	0
149.2 – 153.1	1
153.1 – 157	4

Fig 2.

word_count · Mirrors language_count almost exactly; check whether the two are effectively duplicates.

Show data table

Histogram bins for word_count (median: 2.0).
bin	count
1 – 4.9	3848
4.9 – 8.8	484
8.8 – 12.7	194
12.7 – 16.6	96
16.6 – 20.5	90
20.5 – 24.4	107
24.4 – 28.3	32
28.3 – 32.2	25
32.2 – 36.1	15
36.1 – 40	4
40 – 43.9	8
43.9 – 47.8	8
47.8 – 51.7	6
51.7 – 55.6	5
55.6 – 59.5	3
59.5 – 63.4	3
63.4 – 67.3	8
67.3 – 71.2	3
71.2 – 75.1	3
75.1 – 79	3
79 – 82.9	1
82.9 – 86.8	1
86.8 – 90.7	2
90.7 – 94.6	4
94.6 – 98.5	5
98.5 – 102.4	2
102.4 – 106.3	3
106.3 – 110.2	2
110.2 – 114.1	2
114.1 – 118	3
118 – 121.9	0
121.9 – 125.8	1
125.8 – 129.7	1
129.7 – 133.6	0
133.6 – 137.5	2
137.5 – 141.4	2
141.4 – 145.3	0
145.3 – 149.2	0
149.2 – 153.1	1
153.1 – 157	4

Fig 3.

source_dataset · Confirms every row comes from a single source ('iecor'), so no cross-source comparison is possible.

Show data table

Top values for source_dataset (1 unique shown, of 1 total).
value	count	share
iecor	4981	100.0%

Fig 4.

cognate_id · ID strings cluster tightly at length 10, useful as a sanity check on identifier formatting.

Show data table

Character-length distribution for cognate_id (mean: 9.883557518570568).
chars	count
7 – 7	5
7 – 7	0
7 – 7	0
7 – 7	0
7 – 7	0
7 – 7	0
7 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	44
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	477
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	4455

Fig 5.

Per-column null rate across the corpus. Columns are ordered by input position.

Show data table

Per-column null rate across the corpus.
column	kind	null %
cognate_id	text	0.0%
concept	categorical	0.0%
words	unknown	0.0%
source_dataset	categorical	0.0%
confidence	numeric	0.0%
word_count	numeric	0.0%
language_count	numeric	0.0%
sources	unknown	0.0%

Fig 6.

Pearson correlation across numeric columns (sampled, bounded).

Show data table

Pearson correlation across 3 numeric columns (values clipped to 2 decimals).
	confidence	word_count	language_count
confidence	+nan	+nan	+nan
word_count	+nan	+1.00	+1.00
language_count	+nan	+1.00	+1.00

cognate_id text identifier

This is a primary identifier column: every one of the 4981 values is unique, non-null, single-token, and follows an `iecor:` pattern (length 7-10 chars). With n_unique == n and duplicate_rate 0, it functions as a cognate-set key from the IECoR resource rather than a modelling feature.

Treatment: Use as the join key to cognate metadata; exclude from any model features.

anthropic:claude-opus-4-7 · confidence high

Out[12]:

saturn.columns["cognate_id"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	4,981
len_min	7
len_max	10
len_mean	9.884
len_median	10
len_p95	10
word_mean	1
word_median	1
n_empty	0
n_duplicates	0
duplicate_rate	0
vocab_size	4,981
readability_flesch_mean	121.2
emoji_rate	0
url_rate	0
one_word_rate	1
allcaps_rate	0
boilerplate_rate	0
alert: near_unique	100.0% of rows are unique strings
alert: one_word	100.0% rows are a single word
alert: short_text	95th-percentile length under 20 chars

Fig 7.

Character-length distribution for cognate_id.

Show data table

Character-length distribution for cognate_id (mean: 9.883557518570568).
chars	count
7 – 7	5
7 – 7	0
7 – 7	0
7 – 7	0
7 – 7	0
7 – 7	0
7 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	44
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 8	0
8 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	477
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 9	0
9 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	0
10 – 10	4455

concept categorical other

The 'concept' column is a categorical field that is entirely constant: all 4981 rows hold the same empty-string value, giving cardinality 1 and entropy 0. It carries no information and was flagged for imbalance with a top_rate of 1.0.

Treatment: Drop; the column is constant and contributes no signal.

anthropic:claude-opus-4-7 · confidence high

Out[15]:

saturn.columns["concept"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	1
top_value
top_rate	1
cardinality	1
entropy	0
entropy_ratio	0
alert: imbalance	top value is 100.0% of rows

Fig 8.

Top values for concept.

Show data table

Top values for concept (1 unique shown, of 1 total).
value	count	share
	4981	100.0%

words unknown other

The column is named "words" but saturn skipped profiling it, so its kind is unknown and no descriptive statistics were computed. The only confirmed signals are 4981 non-null rows with a null rate of 0.0; uniqueness, type, and value distribution are all unavailable.

Treatment: Re-profile or manually inspect this column before any downstream use, since saturn skipped it.

anthropic:claude-opus-4-7 · confidence low

Out[18]:

saturn.columns["words"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	—
alert: skipped	no profiler for kind=unknown

source_dataset categorical metadata

This column records the originating dataset for each row, but every one of the 4981 records carries the single value "iecor". Cardinality is 1 and entropy is 0, so the field carries no information for modelling or grouping.

Treatment: Drop; constant column with no discriminative signal.

anthropic:claude-opus-4-7 · confidence high

Out[20]:

saturn.columns["source_dataset"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	1
top_value	iecor
top_rate	1
cardinality	1
entropy	0
entropy_ratio	0
alert: imbalance	top value is 100.0% of rows

Fig 9.

Top values for source_dataset.

Show data table

Top values for source_dataset (1 unique shown, of 1 total).
value	count	share
iecor	4981	100.0%

confidence numeric metadata

The column 'confidence' is a numeric field that is entirely constant: all 4981 rows hold the value 1.0, with zero standard deviation and a single unique value. It carries no information for downstream modelling and likely reflects a default or hard-coded score rather than a measured probability.

Treatment: drop, constant column with no variance

anthropic:claude-opus-4-7 · confidence high

Out[23]:

saturn.columns["confidence"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	1
min	1
max	1
mean	1
median	1
std	0
q1	1
q3	1
iqr	0
skew	0
kurtosis	0
n_outliers	0
outlier_rate	0
zero_rate	0
alert: constant	only one distinct value

Fig 10.

Distribution of confidence. Vertical dash marks the median.

Show data table

Histogram bins for confidence (median: 1.0).
bin	count
0.5 – 0.525	0
0.525 – 0.55	0
0.55 – 0.575	0
0.575 – 0.6	0
0.6 – 0.625	0
0.625 – 0.65	0
0.65 – 0.675	0
0.675 – 0.7	0
0.7 – 0.725	0
0.725 – 0.75	0
0.75 – 0.775	0
0.775 – 0.8	0
0.8 – 0.825	0
0.825 – 0.85	0
0.85 – 0.875	0
0.875 – 0.9	0
0.9 – 0.925	0
0.925 – 0.95	0
0.95 – 0.975	0
0.975 – 1	0
1 – 1.025	4981
1.025 – 1.05	0
1.05 – 1.075	0
1.075 – 1.1	0
1.1 – 1.125	0
1.125 – 1.15	0
1.15 – 1.175	0
1.175 – 1.2	0
1.2 – 1.225	0
1.225 – 1.25	0
1.25 – 1.275	0
1.275 – 1.3	0
1.3 – 1.325	0
1.325 – 1.35	0
1.35 – 1.375	0
1.375 – 1.4	0
1.4 – 1.425	0
1.425 – 1.45	0
1.45 – 1.475	0
1.475 – 1.5	0

word_count numeric feature

Counts of words per record, ranging from 1 to 157 with a median of 2 and IQR of 3. The distribution is severely right-skewed (skew 6.84, kurtosis 59.74) with 649 outliers (13.0% of rows) pulling the mean to 5.17 against std 12.13. Most entries are very short while a long tail of verbose records dominates the variance.

Treatment: Apply a log1p transform before modelling to tame the heavy right tail.

anthropic:claude-opus-4-7 · confidence high

Out[26]:

saturn.columns["word_count"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	93
min	1
max	157
mean	5.168
median	2
std	12.13
q1	1
q3	4
iqr	3
skew	6.837
kurtosis	59.74
n_outliers	649
outlier_rate	0.1303
zero_rate	0
alert: high_skew	skew=+6.84
alert: outliers	13.0% rows beyond 1.5 IQR

Fig 11.

Distribution of word_count. Vertical dash marks the median.

Show data table

Histogram bins for word_count (median: 2.0).
bin	count
1 – 4.9	3848
4.9 – 8.8	484
8.8 – 12.7	194
12.7 – 16.6	96
16.6 – 20.5	90
20.5 – 24.4	107
24.4 – 28.3	32
28.3 – 32.2	25
32.2 – 36.1	15
36.1 – 40	4
40 – 43.9	8
43.9 – 47.8	8
47.8 – 51.7	6
51.7 – 55.6	5
55.6 – 59.5	3
59.5 – 63.4	3
63.4 – 67.3	8
67.3 – 71.2	3
71.2 – 75.1	3
75.1 – 79	3
79 – 82.9	1
82.9 – 86.8	1
86.8 – 90.7	2
90.7 – 94.6	4
94.6 – 98.5	5
98.5 – 102.4	2
102.4 – 106.3	3
106.3 – 110.2	2
110.2 – 114.1	2
114.1 – 118	3
118 – 121.9	0
121.9 – 125.8	1
125.8 – 129.7	1
129.7 – 133.6	0
133.6 – 137.5	2
137.5 – 141.4	2
141.4 – 145.3	0
145.3 – 149.2	0
149.2 – 153.1	1
153.1 – 157	4

language_count numeric feature

A count of languages per record, ranging from 1 to 157 with a median of just 2 and IQR of 3. The distribution is severely right-skewed (skew 6.84, kurtosis 59.77) with 649 outliers (13.0%), meaning a small number of records list dozens of languages while most list only a handful.

Treatment: Log-transform or cap before modelling to tame the heavy tail.

anthropic:claude-opus-4-7 · confidence high

Out[29]:

saturn.columns["language_count"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	94
min	1
max	157
mean	5.166
median	2
std	12.13
q1	1
q3	4
iqr	3
skew	6.838
kurtosis	59.77
n_outliers	649
outlier_rate	0.1303
zero_rate	0
alert: high_skew	skew=+6.84
alert: outliers	13.0% rows beyond 1.5 IQR

Fig 12.

Distribution of language_count. Vertical dash marks the median.

Show data table

Histogram bins for language_count (median: 2.0).
bin	count
1 – 4.9	3849
4.9 – 8.8	483
8.8 – 12.7	194
12.7 – 16.6	96
16.6 – 20.5	90
20.5 – 24.4	107
24.4 – 28.3	32
28.3 – 32.2	25
32.2 – 36.1	15
36.1 – 40	4
40 – 43.9	8
43.9 – 47.8	8
47.8 – 51.7	6
51.7 – 55.6	5
55.6 – 59.5	3
59.5 – 63.4	3
63.4 – 67.3	8
67.3 – 71.2	3
71.2 – 75.1	3
75.1 – 79	4
79 – 82.9	0
82.9 – 86.8	1
86.8 – 90.7	2
90.7 – 94.6	4
94.6 – 98.5	5
98.5 – 102.4	2
102.4 – 106.3	3
106.3 – 110.2	2
110.2 – 114.1	2
114.1 – 118	3
118 – 121.9	0
121.9 – 125.8	1
125.8 – 129.7	1
129.7 – 133.6	0
133.6 – 137.5	2
137.5 – 141.4	2
141.4 – 145.3	0
145.3 – 149.2	0
149.2 – 153.1	1
153.1 – 157	4

sources unknown other

The column 'sources' was skipped by the profiler, so its kind is unknown and no descriptive statistics are available. We can only confirm it has 4981 rows with a null rate of 0.0 and no recorded unique count. Without further evidence, the content and structure cannot be characterised.

Treatment: Re-profile or manually inspect this column before deciding on downstream handling.

anthropic:claude-opus-4-7 · confidence low

Out[32]:

saturn.columns["sources"].stats

stat	value
n	4,981
nulls	0 (0.0%)
unique	—
alert: skipped	no profiler for kind=unknown

How to cite

click to copy

BibTeX

@misc{saturn-processed-cognate-sets-2026,
  author       = {Steuber, Luke},
  title        = {Saturn reading: processed cognate sets},
  year         ={2026},
  howpublished = {\url{https://dr.eamer.dev/saturn/view/processed-cognate_sets}},
  note         = {Profiled with saturn-dissect v0.2.0, prompt saturn-insight-v2, model anthropic:claude-opus-4-7},
}

APA

Steuber, L. (2026). Saturn reading: processed cognate sets. Source: /home/coolhand/servers/diachronica/etymology_atlas/processed/cognate_sets.json. Profiled with saturn-dissect v0.2.0 (saturn-insight-v2, anthropic:claude-opus-4-7). Retrieved from https://dr.eamer.dev/saturn/view/processed-cognate_sets