saturn·

← all readings

quirky silence data

source /home/coolhand/html/datavis/data_trove/data/quirky/silence_data.json 6,998 rows 6 columns profiled 2026-05-01 raw JSON static .html .ipynb Report Notebook

Reading

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

This dataset catalogs 6,998 world languages, each with a name (n), speaker population (p), geographic coordinates (lat/lng), a numeric score (s), and an endangerment status (ss). The most striking feature is the extreme skew in speaker population: the median language has just 11,000 speakers but the mean is over 1.1 million, with a max near 965 million and roughly 13% of entries showing zero speakers — a tell-tale signature of dying or dormant languages. The endangerment status field is also worth a close look: only about 44% of languages are 'safe', while the remaining categories span from 'vulnerable' all the way to 'extinct' (219 cases). Geography is broadly distributed (latitude centered near the tropics, longitude spanning the globe), so the dataset supports both statistical and map-based exploration.

citing: row_count · columns.p.stats.median · columns.p.stats.mean · columns.p.stats.max · columns.p.stats.zero_rate · columns.p.stats.skew · columns.ss.top_rate · columns.ss.top_values · columns.lat.stats.median · columns.lng.stats.median · columns.s.n_unique

ss · How many languages fall into each endangerment status — note that less than half are 'safe'.
Show data table
valuecountshare
safe307443.9%
definitely endangered175325.1%
vulnerable116016.6%
severely endangered3745.3%
critically endangered3274.7%
extinct2193.1%
unknown911.3%
p · Speaker population is extremely right-skewed; expect a long tail and a large spike at zero.
Show data table
bincount
0 – 2.411e+076943
2.411e+07 – 4.823e+0728
4.823e+07 – 7.234e+075
7.234e+07 – 9.646e+0712
9.646e+07 – 1.206e+080
1.206e+08 – 1.447e+083
1.447e+08 – 1.688e+081
1.688e+08 – 1.929e+080
1.929e+08 – 2.17e+080
2.17e+08 – 2.411e+082
2.411e+08 – 2.653e+080
2.653e+08 – 2.894e+080
2.894e+08 – 3.135e+080
3.135e+08 – 3.376e+080
3.376e+08 – 3.617e+080
3.617e+08 – 3.858e+081
3.858e+08 – 4.099e+080
4.099e+08 – 4.34e+080
4.34e+08 – 4.582e+081
4.582e+08 – 4.823e+080
4.823e+08 – 5.064e+080
5.064e+08 – 5.305e+080
5.305e+08 – 5.546e+080
5.546e+08 – 5.787e+080
5.787e+08 – 6.028e+080
6.028e+08 – 6.27e+080
6.27e+08 – 6.511e+080
6.511e+08 – 6.752e+080
6.752e+08 – 6.993e+080
6.993e+08 – 7.234e+081
7.234e+08 – 7.475e+080
7.475e+08 – 7.716e+080
7.716e+08 – 7.958e+080
7.958e+08 – 8.199e+080
8.199e+08 – 8.44e+080
8.44e+08 – 8.681e+080
8.681e+08 – 8.922e+080
8.922e+08 – 9.163e+080
9.163e+08 – 9.404e+080
9.404e+08 – 9.646e+081
lat · Latitude distribution shows where in the world languages cluster, weighted toward the tropics and northern hemisphere.
Show data table
bincount
-55.27 – -52.061
-52.06 – -48.851
-48.85 – -45.641
-45.64 – -42.430
-42.43 – -39.222
-39.22 – -36.015
-36.01 – -32.89
-32.8 – -29.5913
-29.59 – -26.3823
-26.38 – -23.1748
-23.17 – -19.96100
-19.96 – -16.75101
-16.75 – -13.54217
-13.54 – -10.33202
-10.33 – -7.116461
-7.116 – -3.906749
-3.906 – -0.6958640
-0.6958 – 2.514344
2.514 – 5.725434
5.725 – 8.935609
8.935 – 12.15676
12.15 – 15.36264
15.36 – 18.57368
18.57 – 21.78210
21.78 – 24.99282
24.99 – 28.2320
28.2 – 31.41131
31.41 – 34.62124
34.62 – 37.83148
37.83 – 41.0481
41.04 – 44.25107
44.25 – 47.4665
47.46 – 50.6768
50.67 – 53.8868
53.88 – 57.0935
57.09 – 60.320
60.3 – 63.5134
63.51 – 66.7218
66.72 – 69.9313
69.93 – 73.146
s · The 's' score takes only six integer values from 0 to 5; check which level dominates.
Show data table
bincount
0 – 0.125250
0.125 – 0.250
0.25 – 0.3750
0.375 – 0.50
0.5 – 0.6250
0.625 – 0.750
0.75 – 0.8750
0.875 – 10
1 – 1.125328
1.125 – 1.250
1.25 – 1.3750
1.375 – 1.50
1.5 – 1.6250
1.625 – 1.750
1.75 – 1.8750
1.875 – 20
2 – 2.125383
2.125 – 2.250
2.25 – 2.3750
2.375 – 2.50
2.5 – 2.6250
2.625 – 2.750
2.75 – 2.8750
2.875 – 30
3 – 3.1251768
3.125 – 3.250
3.25 – 3.3750
3.375 – 3.50
3.5 – 3.6250
3.625 – 3.750
3.75 – 3.8750
3.875 – 40
4 – 4.1251176
4.125 – 4.250
4.25 – 4.3750
4.375 – 4.50
4.5 – 4.6250
4.625 – 4.750
4.75 – 4.8750
4.875 – 53093
lng · Longitude spread reveals the global reach of the dataset across continents.
Show data table
bincount
-178.8 – -169.811
-169.8 – -160.94
-160.9 – -151.99
-151.9 – -14311
-143 – -13410
-134 – -125.115
-125.1 – -116.194
-116.1 – -107.237
-107.2 – -98.2171
-98.21 – -89.26257
-89.26 – -80.3150
-80.31 – -71.35176
-71.35 – -62.4150
-62.4 – -53.45112
-53.45 – -44.548
-44.5 – -35.5421
-35.54 – -26.590
-26.59 – -17.643
-17.64 – -8.68795
-8.687 – 0.265261
0.265 – 9.217420
9.217 – 18.17687
18.17 – 27.12297
27.12 – 36.07402
36.07 – 45.03211
45.03 – 53.98113
53.98 – 62.9327
62.93 – 71.8873
71.88 – 80.84212
80.84 – 89.79191
89.79 – 98.74242
98.74 – 107.7386
107.7 – 116.6202
116.6 – 125.6437
125.6 – 134.5266
134.5 – 143.5512
143.5 – 152.5597
152.5 – 161.4106
161.4 – 170.4170
170.4 – 179.312

Schema

6 columns
Alerts
n text 0.0% 6,998
near_unique one_word
lat numeric 0.0% 4,048
lng numeric 0.0% 5,560
p numeric 0.0% 1,627
high_skew outliers
s numeric 0.0% 6
ss categorical 0.0% 7

n

text identifier near_unique one_word
Column 'n' holds short text labels, one per row, with all 6998 values unique and no nulls. About 73% are single words (word_mean 1.37, len_mean ~9), and the top tokens — 'language', 'sign', 'zapotec', 'mixtec', 'naga' plus directional modifiers — strongly suggest these are language names (e.g., 'Southern Zapotec', 'sign language' variants). Vocab_size (7003) slightly exceeds n, consistent with a small set of recurring qualifiers across otherwise unique names. Treatment: Treat as a name key; left-join on this rather than using as a model feature. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
6,998
len_min
1
len_max
43
len_mean
8.972
len_median
7
len_p95
21
word_mean
1.369
word_median
1
n_empty
0
n_duplicates
0
duplicate_rate
0
vocab_size
7,003
readability_flesch_mean
57.13
emoji_rate
0
url_rate
0
one_word_rate
0.7314
allcaps_rate
0
boilerplate_rate
0

lat

numeric feature
Latitude coordinates spanning -55.27 to 73.14, covering most inhabited latitudes from sub-Antarctic to high Arctic. The distribution leans north of the equator (mean 8.44, median 6.37) with mild positive skew (0.70) and 149 outliers (2.1%) likely representing extreme polar observations. With 4048 unique values across 6998 rows and no nulls, this is clean geospatial data ready for use. Treatment: Pair with longitude for geospatial features; consider binning by hemisphere or climate zone. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
4,048
min
-55.27
max
73.14
mean
8.437
median
6.37
std
18
q1
-4.65
q3
18.29
iqr
22.94
skew
0.6975
kurtosis
0.4773
n_outliers
149
outlier_rate
0.02129
zero_rate
0

lng

numeric feature
This is a longitude feature spanning -178.78 to 179.31 across 6998 rows with 5560 unique values and no nulls. The distribution is mildly left-skewed (-0.498) with a median of 47.65 and IQR from 8.28 to 123.94, suggesting an Eastern-Hemisphere bias rather than a uniform global spread. Only 12 outliers (0.17%) flagged, which is expected given the bounded geographic range. Treatment: Pair with latitude as a geospatial coordinate; consider cyclical encoding or geohashing rather than treating as a plain scalar. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
5,560
min
-178.8
max
179.3
mean
52.46
median
47.65
std
79.67
q1
8.282
q3
123.9
iqr
115.7
skew
-0.4982
kurtosis
-0.673
n_outliers
12
outlier_rate
0.001715
zero_rate
0

p

numeric feature high_skew outliers
Column `p` is a heavily right-skewed numeric value spanning 0 to 964,553,200 with a median of just 11,000 — likely a monetary amount, count, or size metric. The skew of 39.45 and kurtosis of 1870 are extreme, with 16.56% of rows flagged as outliers and 13.46% sitting at exactly zero. The mean (1,157,741) sits more than 100x above the median, so summary averages will be misleading. Treatment: Apply a log1p transform before modelling and consider a separate zero-vs-nonzero indicator. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
1,627
min
0
max
9.646e+08
mean
1.158e+06
median
11,000
std
1.73e+07
q1
1,200
q3
83,000
iqr
81,800
skew
39.45
kurtosis
1870
n_outliers
1,159
outlier_rate
0.1656
zero_rate
0.1346

s

numeric feature
Column 's' holds an integer-valued score on a 0-5 scale with only 6 unique values across 6998 rows and no nulls. The distribution skews high (mean 3.80, median 4.0, skew -1.04), with Q1-Q3 spanning 3 to 5 and about 3.6% of rows at zero. This looks like an ordinal rating rather than a continuous measurement. Treatment: Treat as an ordinal rating; consider ordinal encoding or bucket low scores before modelling. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
6
min
0
max
5
mean
3.796
median
4
std
1.366
q1
3
q3
5
iqr
2
skew
-1.041
kurtosis
0.4154
n_outliers
0
outlier_rate
0
zero_rate
0.03572

ss

categorical label
The column 'ss' is a categorical safety/endangerment status with 7 distinct levels ranging from 'safe' to 'extinct', plus an 'unknown' bucket. 'safe' dominates at 43.9% (3074 of 6998) with 'definitely endangered' second at 1753, and there are no nulls. The distribution is moderately balanced (entropy ratio 0.76) but the tail categories like 'extinct' (219) and 'unknown' (91) are sparse. Treatment: Encode as an ordinal scale (safe → extinct) and treat 'unknown' separately or impute. high · anthropic:claude-opus-4-7
n
6,998
nulls
0 (0.0%)
unique
7
top_value
safe
top_rate
0.4393
cardinality
7
entropy
2.122
entropy_ratio
0.7557