usgs-significant-earthquakes-usgs_significant_earthquakes

Overview

Source: /home/coolhand/datasets/usgs-significant-earthquakes/usgs_significant_earthquakes.json

Saturn profiled 3,742 rows across 11 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/datasets/usgs-significant-earthquakes/usgs_significant_earthquakes.json",
    "--findings", "usgs-significant-earthquakes-usgs_significant_earthquakes.json",
    "--llm", "anthropic:claude-opus-4-7",
])

Summary confidence: high

This dataset contains 3,742 records of significant earthquakes from USGS, with 11 columns covering location (latitude, longitude, place/name), magnitude, depth, and event metadata. Magnitude is tightly clustered between 4.5 and 5.1 (median 4.8) but has a long right tail reaching 8.2, with 184 outliers worth examining for the rare large events. Depth_km is highly skewed (skew 3.07) with a median of 10 km but a max of 248.7 km and 314 outliers, suggesting a mix of shallow and deep-focus quakes. Geographically, the data is heavily concentrated around Alaska — 'alaska' appears in 1,991 place names and 'off the coast of Oregon' alone accounts for 151 records — so this is effectively a North Pacific / Alaska-dominated sample rather than a global one. Note that the `category` column is constant ('significant_earthquakes') and `earthquake_type` is 99.9% 'earthquake', so neither will be useful for segmentation.

citing: row_count · column_count · depth_km · magnitude · latitude · longitude · name · place · earthquake_type · category

Out[4]:

saturn.schema() · 11 columns

column	kind	n	null%	unique	alerts
latitude	numeric	3,742	0.0%	3,627
longitude	numeric	3,742	0.0%	3,668
name	text	3,742	0.0%	3,002	multilingual
description	text	3,742	0.0%	3,591	near_unique
category	categorical	3,742	0.0%	1	imbalance
date	text	3,742	0.0%	3,741	near_unique one_word allcaps short_text
country	unknown	3,742	0.0%	—	skipped
magnitude	numeric	3,742	0.0%	123
depth_km	numeric	3,742	0.0%	1,505	high_skew outliers
place	text	3,742	0.0%	3,002	multilingual
earthquake_type	categorical	3,742	0.0%	3	imbalance

Fig 1.

magnitude · Most events fall between 4.5 and 5.1; watch the long right tail toward magnitude 8.2.

Show data table

Histogram bins for magnitude (median: 4.8).
bin	count
4.5 – 4.593	752
4.593 – 4.685	601
4.685 – 4.777	445
4.777 – 4.87	340
4.87 – 4.963	286
4.963 – 5.055	254
5.055 – 5.147	218
5.147 – 5.24	177
5.24 – 5.332	137
5.332 – 5.425	104
5.425 – 5.518	85
5.518 – 5.61	66
5.61 – 5.702	53
5.702 – 5.795	3
5.795 – 5.887	38
5.887 – 5.98	46
5.98 – 6.072	25
6.072 – 6.165	17
6.165 – 6.257	16
6.257 – 6.35	11
6.35 – 6.442	15
6.442 – 6.535	11
6.535 – 6.627	10
6.627 – 6.72	4
6.72 – 6.812	6
6.812 – 6.905	5
6.905 – 6.997	0
6.997 – 7.09	3
7.09 – 7.182	3
7.182 – 7.275	3
7.275 – 7.367	1
7.367 – 7.46	0
7.46 – 7.552	1
7.552 – 7.645	1
7.645 – 7.737	0
7.737 – 7.83	2
7.83 – 7.922	2
7.922 – 8.015	0
8.015 – 8.107	0
8.107 – 8.2	1

Fig 2.

depth_km · Highly skewed — most quakes are shallow near 10 km, but 314 outliers extend to 248 km deep.

Show data table

Histogram bins for depth_km (median: 10.0).
bin	count
-2.261 – 4.013	219
4.013 – 10.29	1730
10.29 – 16.56	370
16.56 – 22.84	258
22.84 – 29.11	230
29.11 – 35.38	250
35.38 – 41.66	167
41.66 – 47.93	129
47.93 – 54.21	56
54.21 – 60.48	31
60.48 – 66.75	43
66.75 – 73.03	27
73.03 – 79.3	19
79.3 – 85.58	29
85.58 – 91.85	21
91.85 – 98.12	19
98.12 – 104.4	24
104.4 – 110.7	12
110.7 – 116.9	9
116.9 – 123.2	14
123.2 – 129.5	13
129.5 – 135.8	19
135.8 – 142	14
142 – 148.3	5
148.3 – 154.6	6
154.6 – 160.9	0
160.9 – 167.1	7
167.1 – 173.4	4
173.4 – 179.7	0
179.7 – 186	1
186 – 192.2	5
192.2 – 198.5	1
198.5 – 204.8	3
204.8 – 211.1	2
211.1 – 217.3	3
217.3 – 223.6	1
223.6 – 229.9	0
229.9 – 236.2	0
236.2 – 242.4	0
242.4 – 248.7	1

Fig 3.

latitude · Latitudes cluster between 41° and 56° N, confirming a northern-hemisphere, Alaska-heavy sample.

Show data table

Histogram bins for latitude (median: 52.395700000000005).
bin	count
20.02 – 21.26	35
21.26 – 22.51	28
22.51 – 23.75	42
23.75 – 25	88
25 – 26.24	91
26.24 – 27.49	35
27.49 – 28.73	43
28.73 – 29.98	39
29.98 – 31.22	52
31.22 – 32.46	83
32.46 – 33.71	52
33.71 – 34.95	26
34.95 – 36.2	67
36.2 – 37.44	43
37.44 – 38.69	54
38.69 – 39.93	22
39.93 – 41.18	131
41.18 – 42.42	52
42.42 – 43.66	96
43.66 – 44.91	177
44.91 – 46.15	4
46.15 – 47.4	7
47.4 – 48.64	34
48.64 – 49.89	117
49.89 – 51.13	151
51.13 – 52.38	288
52.38 – 53.62	423
53.62 – 54.86	349
54.86 – 56.11	223
56.11 – 57.35	201
57.35 – 58.6	75
58.6 – 59.84	104
59.84 – 61.09	116
61.09 – 62.33	117
62.33 – 63.58	120
63.58 – 64.82	33
64.82 – 66.06	35
66.06 – 67.31	25
67.31 – 68.55	29
68.55 – 69.8	35

Fig 4.

place · Top places are dominated by Alaska and Oregon coast locations, indicating regional concentration.

Show data table

Character-length distribution for place (mean: 29.465793693212184).
chars	count
4 – 5	1
5 – 7	0
7 – 8	1
8 – 10	0
10 – 11	0
11 – 12	0
12 – 14	2
14 – 15	11
15 – 16	40
16 – 18	1
18 – 19	20
19 – 20	19
20 – 22	8
22 – 23	219
23 – 25	60
25 – 26	122
26 – 27	543
27 – 29	499
29 – 30	823
30 – 32	325
32 – 33	362
33 – 34	378
34 – 36	105
36 – 37	40
37 – 38	37
38 – 40	25
40 – 41	34
41 – 42	3
42 – 44	0
44 – 45	15
45 – 47	22
47 – 48	14
48 – 49	3
49 – 51	1
51 – 52	2
52 – 54	5
54 – 55	0
55 – 56	0
56 – 58	0
58 – 59	2

Fig 5.

earthquake_type · Nearly all records (99.9%) are 'earthquake', with only a handful of explosions and one landslide.

Show data table

Top values for earthquake_type (3 unique shown, of 3 total).
value	count	share
earthquake	3739	99.9%
explosion	2	0.1%
landslide	1	0.0%

Fig 6.

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 %
latitude	numeric	0.0%
longitude	numeric	0.0%
name	text	0.0%
description	text	0.0%
category	categorical	0.0%
date	text	0.0%
country	unknown	0.0%
magnitude	numeric	0.0%
depth_km	numeric	0.0%
place	text	0.0%
earthquake_type	categorical	0.0%

Fig 7.

Language mix across all text columns (per-string detection, sampled).

Show data table

Per-language counts (total 7,480 detected strings).
lang	count	share
en	7438	99.4%
es	32	0.4%
de	6	0.1%
ja	2	0.0%
ceb	2	0.0%

Fig 8.

Pearson correlation across numeric columns (sampled, bounded).

Show data table

Pearson correlation across 4 numeric columns (values clipped to 2 decimals).
	latitude	longitude	magnitude	depth_km
latitude	+1.00	-0.72	-0.12	+0.31
longitude	-0.72	+1.00	+0.07	-0.37
magnitude	-0.12	+0.07	+1.00	-0.03
depth_km	+0.31	-0.37	-0.03	+1.00

latitude numeric feature

Geographic latitude coordinate, with values ranging from 20.02 to 69.7975 and a median of 52.40, consistent with locations spanning roughly the tropics to the Arctic Circle. The distribution is left-skewed (skew -0.76), concentrated in northern mid-to-high latitudes (Q1-Q3: 41.34-55.90), suggesting a Europe/North America bias. No nulls or outliers, and 3627 unique values across 3742 rows indicate near-row-level granularity.

Treatment: Pair with longitude for geospatial features; consider binning or clustering rather than treating as a plain scalar.

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

Out[14]:

saturn.columns["latitude"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,627
min	20.02
max	69.8
mean	48.53
median	52.4
std	11.58
q1	41.34
q3	55.9
iqr	14.56
skew	-0.7591
kurtosis	-0.2887
n_outliers	0
outlier_rate	0
zero_rate	0

Fig 9.

Distribution of latitude. Vertical dash marks the median.

Show data table

Histogram bins for latitude (median: 52.395700000000005).
bin	count
20.02 – 21.26	35
21.26 – 22.51	28
22.51 – 23.75	42
23.75 – 25	88
25 – 26.24	91
26.24 – 27.49	35
27.49 – 28.73	43
28.73 – 29.98	39
29.98 – 31.22	52
31.22 – 32.46	83
32.46 – 33.71	52
33.71 – 34.95	26
34.95 – 36.2	67
36.2 – 37.44	43
37.44 – 38.69	54
38.69 – 39.93	22
39.93 – 41.18	131
41.18 – 42.42	52
42.42 – 43.66	96
43.66 – 44.91	177
44.91 – 46.15	4
46.15 – 47.4	7
47.4 – 48.64	34
48.64 – 49.89	117
49.89 – 51.13	151
51.13 – 52.38	288
52.38 – 53.62	423
53.62 – 54.86	349
54.86 – 56.11	223
56.11 – 57.35	201
57.35 – 58.6	75
58.6 – 59.84	104
59.84 – 61.09	116
61.09 – 62.33	117
62.33 – 63.58	120
63.58 – 64.82	33
64.82 – 66.06	35
66.06 – 67.31	25
67.31 – 68.55	29
68.55 – 69.8	35

longitude numeric feature

Geographic longitude coordinates, all in the western hemisphere with values ranging from -169.997 to -65.039 (mean -140.10, median -144.22). The distribution is mildly right-skewed (0.45) with 3668 unique values across 3742 rows, suggesting near-distinct location points. The 26 outliers (0.69%) likely correspond to easternmost points well outside the dense western cluster bounded by Q1=-159.93 and Q3=-125.10.

Treatment: Pair with latitude for geospatial features (e.g., binning, clustering, or distance computation) rather than treating as a standalone scalar.

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

Out[17]:

saturn.columns["longitude"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,668
min	-170
max	-65.04
mean	-140.1
median	-144.2
std	21.81
q1	-159.9
q3	-125.1
iqr	34.84
skew	0.4489
kurtosis	-0.4302
n_outliers	26
outlier_rate	0.006948
zero_rate	0

Fig 10.

Distribution of longitude. Vertical dash marks the median.

Show data table

Histogram bins for longitude (median: -144.21715).
bin	count
-170 – -167.4	368
-167.4 – -164.7	161
-164.7 – -162.1	243
-162.1 – -159.5	241
-159.5 – -156.9	113
-156.9 – -154.3	129
-154.3 – -151.6	241
-151.6 – -149	220
-149 – -146.4	77
-146.4 – -143.8	83
-143.8 – -141.1	29
-141.1 – -138.5	48
-138.5 – -135.9	41
-135.9 – -133.3	34
-133.3 – -130.6	129
-130.6 – -128	428
-128 – -125.4	200
-125.4 – -122.8	89
-122.8 – -120.1	44
-120.1 – -117.5	130
-117.5 – -114.9	149
-114.9 – -112.3	113
-112.3 – -109.6	163
-109.6 – -107	134
-107 – -104.4	33
-104.4 – -101.8	12
-101.8 – -99.15	10
-99.15 – -96.53	20
-96.53 – -93.9	4
-93.9 – -91.28	1
-91.28 – -88.65	2
-88.65 – -86.03	6
-86.03 – -83.41	1
-83.41 – -80.78	2
-80.78 – -78.16	3
-78.16 – -75.53	7
-75.53 – -72.91	8
-72.91 – -70.29	7
-70.29 – -67.66	5
-67.66 – -65.04	14

name text metadata

This 'name' column reads as human-readable seismic event location descriptions (e.g., '104 km SSW of Nikolski, Alaska'), dominated by distance-and-bearing phrases — 'km', 'of', and cardinal directions like 'SSW'/'SSE' top the word list. Of 3742 rows there are only 3002 uniques and a 19.8% duplicate rate, with 'off the coast of Oregon' appearing 151 times, suggesting many events share the same place label. Despite a 'multilingual' flag, 3719 of 3742 strings are detected as English with only a handful tagged de/es/ja/ceb — likely false positives on short toponyms rather than real language mixing.

Treatment: Parse into structured fields (distance_km, bearing, place) rather than embedding the raw string.

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

Out[20]:

saturn.columns["name"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,002
len_min	4
len_max	59
len_mean	29.47
len_median	29
len_p95	36
word_mean	6.293
word_median	6
n_empty	0
n_duplicates	740
duplicate_rate	0.1978
vocab_size	1,036
readability_flesch_mean	69.91
emoji_rate	0
url_rate	0
one_word_rate	0.0005345
allcaps_rate	0
boilerplate_rate	0
alert: multilingual	6 languages detected in sample

Fig 11.

Character-length distribution for name.

Show data table

Character-length distribution for name (mean: 29.465793693212184).
chars	count
4 – 5	1
5 – 7	0
7 – 8	1
8 – 10	0
10 – 11	0
11 – 12	0
12 – 14	2
14 – 15	11
15 – 16	40
16 – 18	1
18 – 19	20
19 – 20	19
20 – 22	8
22 – 23	219
23 – 25	60
25 – 26	122
26 – 27	543
27 – 29	499
29 – 30	823
30 – 32	325
32 – 33	362
33 – 34	378
34 – 36	105
36 – 37	40
37 – 38	37
38 – 40	25
40 – 41	34
41 – 42	3
42 – 44	0
44 – 45	15
45 – 47	22
47 – 48	14
48 – 49	3
49 – 51	1
51 – 52	2
52 – 54	5
54 – 55	0
55 – 56	0
56 – 58	0
58 – 59	2

description text free_text

Templated earthquake event descriptions, e.g. magnitude/depth strings ending in locations like Alaska. Lengths are tightly clustered (min 45, max 100, p95 79) and every row contains 'magnitude', '-', and 'depth:', confirming a fixed format. Despite that template, 3591 of 3742 values are unique with only 151 duplicates (4.0%), so the free-text portion (location, magnitude, depth) carries the signal.

Treatment: Parse out magnitude, depth, and location with a regex rather than embedding the templated string.

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

Out[23]:

saturn.columns["description"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,591
len_min	45
len_max	100
len_mean	71.71
len_median	72
len_p95	79
word_mean	12.29
word_median	12
n_empty	0
n_duplicates	151
duplicate_rate	0.04035
vocab_size	2,674
readability_flesch_mean	63.23
emoji_rate	0
url_rate	0
one_word_rate	0
allcaps_rate	0
boilerplate_rate	0
alert: near_unique	96.0% of rows are unique strings

Fig 12.

Character-length distribution for description.

Show data table

Character-length distribution for description (mean: 71.7116515232496).
chars	count
45 – 46	1
46 – 48	1
48 – 49	0
49 – 50	0
50 – 52	0
52 – 53	0
53 – 55	4
55 – 56	4
56 – 57	19
57 – 59	15
59 – 60	20
60 – 62	9
62 – 63	17
63 – 64	158
64 – 66	45
66 – 67	73
67 – 68	380
68 – 70	350
70 – 71	726
71 – 72	389
72 – 74	367
74 – 75	605
75 – 77	204
77 – 78	91
78 – 79	90
79 – 81	21
81 – 82	33
82 – 84	8
84 – 85	25
85 – 86	31
86 – 88	24
88 – 89	10
89 – 90	7
90 – 92	1
92 – 93	3
93 – 94	5
94 – 96	1
96 – 97	2
97 – 99	1
99 – 100	2

category categorical metadata

This column tags every one of the 3,742 rows with the single value "significant_earthquakes", giving cardinality 1 and entropy 0. It carries no information for modelling and most likely encodes the dataset's source or slice rather than a per-row attribute.

Treatment: Drop before modelling; retain only as a dataset-level provenance tag.

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

Out[26]:

saturn.columns["category"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	1
top_value	significant_earthquakes
top_rate	1
cardinality	1
entropy	0
entropy_ratio	0
alert: imbalance	top value is 100.0% of rows

Fig 13.

Top values for category.

Show data table

Top values for category (1 unique shown, of 1 total).
value	count	share
significant_earthquakes	3742	100.0%

date text identifier

Despite the name, this column holds 18-19 character single-token strings shaped like '-01-01' rather than parseable dates — every value ends in '-01-01' and the prefix appears to be a 13-digit epoch-style integer. With 3741 unique values across 3742 rows (one duplicate) and a 100% one-word, all-caps rate, it behaves as a near-unique identifier, not a temporal feature.

Treatment: Drop as-is or parse the leading numeric prefix into a real timestamp before any time-based use.

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

Out[29]:

saturn.columns["date"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,741
len_min	18
len_max	19
len_mean	18.95
len_median	19
len_p95	19
word_mean	1
word_median	1
n_empty	0
n_duplicates	1
duplicate_rate	0.0002672
vocab_size	3,741
readability_flesch_mean	121.2
emoji_rate	0
url_rate	0
one_word_rate	1
allcaps_rate	1
boilerplate_rate	0
alert: near_unique	100.0% of rows are unique strings
alert: one_word	100.0% rows are a single word
alert: allcaps	100.0% rows are all-caps
alert: short_text	95th-percentile length under 20 chars

Fig 14.

Character-length distribution for date.

Show data table

Character-length distribution for date (mean: 18.951897381079636).
chars	count
18 – 18	180
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 18	0
18 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	0
19 – 19	3562

country unknown metadata

This column is labelled "country" and has 3742 fully populated rows with no nulls, but saturn skipped detailed profiling so kind, uniqueness, and value distribution are unknown. Without cardinality or sample values, it is impossible to confirm whether entries are ISO codes, full names, or mixed representations. Treat the absence of stats as the main signal here.

Treatment: Re-profile with type detection enabled to confirm cardinality and standardise country representation before use.

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

Out[32]:

saturn.columns["country"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	—
alert: skipped	no profiler for kind=unknown

magnitude numeric feature

This is a numeric magnitude field, almost certainly earthquake magnitudes given the floor at 4.5 and ceiling at 8.2. Values cluster tightly (median 4.8, IQR 0.5) but the distribution is heavily right-skewed (skew 1.97, kurtosis 5.58) with 184 outliers (4.9%) in the upper tail. Only 123 unique values across 3742 rows suggests reporting at one-decimal precision.

Treatment: Log-transform or bucket into magnitude bands before modelling to tame the right skew.

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

Out[34]:

saturn.columns["magnitude"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	123
min	4.5
max	8.2
mean	4.917
median	4.8
std	0.462
q1	4.6
q3	5.1
iqr	0.5
skew	1.97
kurtosis	5.583
n_outliers	184
outlier_rate	0.04917
zero_rate	0

Fig 15.

Distribution of magnitude. Vertical dash marks the median.

Show data table

Histogram bins for magnitude (median: 4.8).
bin	count
4.5 – 4.593	752
4.593 – 4.685	601
4.685 – 4.777	445
4.777 – 4.87	340
4.87 – 4.963	286
4.963 – 5.055	254
5.055 – 5.147	218
5.147 – 5.24	177
5.24 – 5.332	137
5.332 – 5.425	104
5.425 – 5.518	85
5.518 – 5.61	66
5.61 – 5.702	53
5.702 – 5.795	3
5.795 – 5.887	38
5.887 – 5.98	46
5.98 – 6.072	25
6.072 – 6.165	17
6.165 – 6.257	16
6.257 – 6.35	11
6.35 – 6.442	15
6.442 – 6.535	11
6.535 – 6.627	10
6.627 – 6.72	4
6.72 – 6.812	6
6.812 – 6.905	5
6.905 – 6.997	0
6.997 – 7.09	3
7.09 – 7.182	3
7.182 – 7.275	3
7.275 – 7.367	1
7.367 – 7.46	0
7.46 – 7.552	1
7.552 – 7.645	1
7.645 – 7.737	0
7.737 – 7.83	2
7.83 – 7.922	2
7.922 – 8.015	0
8.015 – 8.107	0
8.107 – 8.2	1

depth_km numeric feature

Numeric depth measurements in kilometers, almost certainly earthquake hypocenter depths given the median of 10.0 and range from -2.261 to 248.7. The distribution is heavily right-skewed (skew 3.07, kurtosis 11.6) with 314 outliers (8.4%) and a Q1 equal to the median at 10.0, suggesting many events are pinned to a default 10 km depth. Negative minimum (-2.261) likely reflects events above sea level or reference datum.

Treatment: Log-transform (after shifting for negatives) and flag the 10 km default-depth pile-up before modelling.

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

Out[37]:

saturn.columns["depth_km"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	1,505
min	-2.261
max	248.7
mean	23.71
median	10
std	28.79
q1	10
q3	29.1
iqr	19.1
skew	3.072
kurtosis	11.61
n_outliers	314
outlier_rate	0.08391
zero_rate	0.002672
alert: high_skew	skew=+3.07
alert: outliers	8.4% rows beyond 1.5 IQR

Fig 16.

Distribution of depth_km. Vertical dash marks the median.

Show data table

Histogram bins for depth_km (median: 10.0).
bin	count
-2.261 – 4.013	219
4.013 – 10.29	1730
10.29 – 16.56	370
16.56 – 22.84	258
22.84 – 29.11	230
29.11 – 35.38	250
35.38 – 41.66	167
41.66 – 47.93	129
47.93 – 54.21	56
54.21 – 60.48	31
60.48 – 66.75	43
66.75 – 73.03	27
73.03 – 79.3	19
79.3 – 85.58	29
85.58 – 91.85	21
91.85 – 98.12	19
98.12 – 104.4	24
104.4 – 110.7	12
110.7 – 116.9	9
116.9 – 123.2	14
123.2 – 129.5	13
129.5 – 135.8	19
135.8 – 142	14
142 – 148.3	5
148.3 – 154.6	6
154.6 – 160.9	0
160.9 – 167.1	7
167.1 – 173.4	4
173.4 – 179.7	0
179.7 – 186	1
186 – 192.2	5
192.2 – 198.5	1
198.5 – 204.8	3
204.8 – 211.1	2
211.1 – 217.3	3
217.3 – 223.6	1
223.6 – 229.9	0
229.9 – 236.2	0
236.2 – 242.4	0
242.4 – 248.7	1

place text metadata

Free-text place descriptions for seismic events, typically formatted as distance + direction + landmark (e.g. '104 km SSW of Nikolski, Alaska'), averaging 29 characters and 6 words. Alaska dominates with 1991 mentions, followed by Canada (472) and Mexico (412), and 'off the coast of Oregon' alone repeats 151 times, driving a 19.8% duplicate rate across 3002 unique values. The multilingual flag is essentially noise: 3719 of 3742 rows are English, with only a handful of de/es/ja/ceb tags likely from place names misread by the detector.

Treatment: Parse into structured fields (distance_km, bearing, reference_place, region) rather than embedding the raw string.

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

Out[40]:

saturn.columns["place"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3,002
len_min	4
len_max	59
len_mean	29.47
len_median	29
len_p95	36
word_mean	6.293
word_median	6
n_empty	0
n_duplicates	740
duplicate_rate	0.1978
vocab_size	1,036
readability_flesch_mean	69.91
emoji_rate	0
url_rate	0
one_word_rate	0.0005345
allcaps_rate	0
boilerplate_rate	0
alert: multilingual	6 languages detected in sample

Fig 17.

Character-length distribution for place.

Show data table

Character-length distribution for place (mean: 29.465793693212184).
chars	count
4 – 5	1
5 – 7	0
7 – 8	1
8 – 10	0
10 – 11	0
11 – 12	0
12 – 14	2
14 – 15	11
15 – 16	40
16 – 18	1
18 – 19	20
19 – 20	19
20 – 22	8
22 – 23	219
23 – 25	60
25 – 26	122
26 – 27	543
27 – 29	499
29 – 30	823
30 – 32	325
32 – 33	362
33 – 34	378
34 – 36	105
36 – 37	40
37 – 38	37
38 – 40	25
40 – 41	34
41 – 42	3
42 – 44	0
44 – 45	15
45 – 47	22
47 – 48	14
48 – 49	3
49 – 51	1
51 – 52	2
52 – 54	5
54 – 55	0
55 – 56	0
56 – 58	0
58 – 59	2

earthquake_type categorical feature

Classifies seismic events into earthquake, explosion, or landslide, but is effectively a constant: 3,739 of 3,742 rows (top_rate 0.9992) are 'earthquake', with only 2 explosions and 1 landslide. Entropy of 0.0101 (entropy_ratio 0.0064) confirms there is virtually no information content here.

Treatment: Drop as a predictor, or isolate the 3 non-earthquake rows as anomalies.

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

Out[43]:

saturn.columns["earthquake_type"].stats

stat	value
n	3,742
nulls	0 (0.0%)
unique	3
top_value	earthquake
top_rate	0.9992
cardinality	3
entropy	0.01014
entropy_ratio	0.006396
alert: imbalance	top value is 99.9% of rows

Fig 18.

Top values for earthquake_type.

Show data table

Top values for earthquake_type (3 unique shown, of 3 total).
value	count	share
earthquake	3739	99.9%
explosion	2	0.1%
landslide	1	0.0%

usgs significant earthquakes usgs significant earthquakes

Overview

Summary confidence: high

latitude numeric feature

longitude numeric feature

name text metadata

description text free_text

category categorical metadata

date text identifier

country unknown metadata

magnitude numeric feature

depth_km numeric feature

place text metadata

earthquake_type categorical feature

How to cite