saturn·

← all readings

environmental desert data

saturn notebook · generated 2026-05-01 Report Notebook

Overview

Source: /home/coolhand/html/datavis/data_trove/data/environmental/desert_data.json

Saturn profiled 52,037 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/html/datavis/data_trove/data/environmental/desert_data.json",
    "--findings", "environmental-desert_data.json",
    "--llm", "anthropic:claude-opus-4-7",
])

Summary confidence: high

This dataset contains 52,037 records describing US Census-tract-level demographics, with an 11-character ID, county and state labels, and five numeric measures: distance/share, income, population, poverty rate, and SNAP counts. State coverage spans all 51 entries (50 states plus DC), led by Texas (4,010), California (3,727), and Florida (3,018), and counties are dominated by common names like Jefferson and Montgomery. The income distribution is right-skewed (mean $78,215 vs median $70,455, max $250,001) with about 4% flagged as outliers, and poverty rate shows a similar skew (mean 13.7%, median 10.8%, max 99.5%). Worth a closer look: the strong skew and outlier rates in inc, pov, and snap, plus how dist_share spreads almost uniformly from 0 to 10,000 (kurtosis -1.5), suggesting it may be a percentile-style metric rather than a raw count.

citing: row_count · column_count · columns.st.top_values · columns.inc.stats · columns.pov.stats · columns.snap.stats · columns.dist_share.stats · columns.cty.top_values

Out[4]:

saturn.schema() · 8 columns

column kind n null% unique alerts
id text 52,037 0.0% 52,037 near_unique one_word allcaps short_text
st categorical 52,037 0.0% 51
cty text 52,037 0.0% 1,870 short_text duplicates
pov numeric 52,037 0.0% 693
inc numeric 52,037 0.0% 31,375
dist_share numeric 52,037 0.0% 33,000
pop numeric 52,037 0.0% 8,732
snap numeric 52,037 0.0% 1,034
Fig 1.
st · State coverage is broad but tilted toward Texas, California, and Florida.
Show data table
valuecountshare
Texas40107.7%
California37277.2%
Florida30185.8%
Ohio23024.4%
Pennsylvania22424.3%
Michigan20734.0%
New York20594.0%
North Carolina19353.7%
Illinois19063.7%
Georgia17513.4%
Virginia14412.8%
Tennessee13582.6%
Indiana13282.6%
New Jersey12102.3%
Missouri11652.2%
Washington11272.2%
Wisconsin10922.1%
Minnesota10852.1%
Alabama10812.1%
Arizona10762.1%
Fig 2.
inc · Income is right-skewed with a long tail up to $250K and ~4% outliers.
Show data table
bincount
2499 – 86874
8687 – 1.487e+0438
1.487e+04 – 2.106e+04193
2.106e+04 – 2.725e+04634
2.725e+04 – 3.344e+041297
3.344e+04 – 3.962e+042050
3.962e+04 – 4.581e+043155
4.581e+04 – 5.2e+044048
5.2e+04 – 5.819e+044797
5.819e+04 – 6.437e+045037
6.437e+04 – 7.056e+044829
7.056e+04 – 7.675e+044317
7.675e+04 – 8.294e+043593
8.294e+04 – 8.912e+042932
8.912e+04 – 9.531e+042448
9.531e+04 – 1.015e+052120
1.015e+05 – 1.077e+051859
1.077e+05 – 1.139e+051463
1.139e+05 – 1.201e+051241
1.201e+05 – 1.262e+051079
1.262e+05 – 1.324e+05910
1.324e+05 – 1.386e+05701
1.386e+05 – 1.448e+05506
1.448e+05 – 1.51e+05484
1.51e+05 – 1.572e+05382
1.572e+05 – 1.634e+05361
1.634e+05 – 1.696e+05261
1.696e+05 – 1.758e+05178
1.758e+05 – 1.819e+05171
1.819e+05 – 1.881e+05131
1.881e+05 – 1.943e+05141
1.943e+05 – 2.005e+05106
2.005e+05 – 2.067e+05100
2.067e+05 – 2.129e+0563
2.129e+05 – 2.191e+0564
2.191e+05 – 2.253e+0562
2.253e+05 – 2.314e+0540
2.314e+05 – 2.376e+0529
2.376e+05 – 2.438e+0532
2.438e+05 – 2.5e+05181
Fig 3.
pov · Poverty rate clusters below 20% but stretches to nearly 100% in extreme tracts.
Show data table
bincount
0 – 2.4872933
2.487 – 4.9757061
4.975 – 7.4627249
7.462 – 9.956599
9.95 – 12.445536
12.44 – 14.924615
14.92 – 17.413752
17.41 – 19.92876
19.9 – 22.392480
22.39 – 24.881844
24.88 – 27.361480
27.36 – 29.851171
29.85 – 32.341029
32.34 – 34.82766
34.82 – 37.31607
37.31 – 39.8488
39.8 – 42.29399
42.29 – 44.77278
44.77 – 47.26222
47.26 – 49.75148
49.75 – 52.24114
52.24 – 54.72112
54.72 – 57.2178
57.21 – 59.748
59.7 – 62.1942
62.19 – 64.6736
64.67 – 67.1625
67.16 – 69.6514
69.65 – 72.1410
72.14 – 74.627
74.62 – 77.116
77.11 – 79.64
79.6 – 82.095
82.09 – 84.572
84.57 – 87.060
87.06 – 89.550
89.55 – 92.040
92.04 – 94.520
94.52 – 97.010
97.01 – 99.51
Fig 4.
dist_share · Distribution is nearly flat from 0 to 10,000 — check whether this is a ranked or percentile measure.
Show data table
bincount
0 – 2505132
250 – 5001829
500 – 7501421
750 – 10001277
1000 – 12501119
1250 – 15001073
1500 – 17501017
1750 – 2000973
2000 – 2250976
2250 – 2500907
2500 – 2750894
2750 – 3000893
3000 – 3250857
3250 – 3500862
3500 – 3750840
3750 – 4000865
4000 – 4250822
4250 – 4500829
4500 – 4750876
4750 – 5000844
5000 – 5250871
5250 – 5500834
5500 – 5750839
5750 – 6000845
6000 – 6250767
6250 – 6500798
6500 – 6750779
6750 – 7000878
7000 – 7250872
7250 – 7500803
7500 – 7750788
7750 – 8000843
8000 – 8250810
8250 – 8500861
8500 – 8750787
8750 – 9000871
9000 – 9250952
9250 – 95001070
9500 – 97501401
9750 – 1e+0411062
Fig 5.
snap · SNAP counts are right-skewed with a median of 146 and a max near 1,900.
Show data table
bincount
0 – 47.29587
47.2 – 94.48464
94.4 – 141.67361
141.6 – 188.86046
188.8 – 2364810
236 – 283.23869
283.2 – 330.42938
330.4 – 377.62328
377.6 – 424.81718
424.8 – 4721286
472 – 519.2980
519.2 – 566.4673
566.4 – 613.6492
613.6 – 660.8417
660.8 – 708288
708 – 755.2213
755.2 – 802.4139
802.4 – 849.6119
849.6 – 896.877
896.8 – 94468
944 – 991.245
991.2 – 103829
1038 – 108622
1086 – 113319
1133 – 11809
1180 – 122710
1227 – 12746
1274 – 13225
1322 – 13697
1369 – 14160
1416 – 14631
1463 – 15102
1510 – 15581
1558 – 16050
1605 – 16521
1652 – 16991
1699 – 17462
1746 – 17942
1794 – 18411
1841 – 18881
Fig 6.
Per-column null rate across the corpus. Columns are ordered by input position.
Show data table
columnkindnull %
idtext0.0%
stcategorical0.0%
ctytext0.0%
povnumeric0.0%
incnumeric0.0%
dist_sharenumeric0.0%
popnumeric0.0%
snapnumeric0.0%
Fig 7.
Pearson correlation across numeric columns (sampled, bounded).
Show data table
povincdist_sharepopsnap
pov+1.00-0.66-0.18-0.21+0.58
inc-0.66+1.00-0.00+0.19-0.56
dist_share-0.18-0.00+1.00-0.01-0.12
pop-0.21+0.19-0.01+1.00+0.28
snap+0.58-0.56-0.12+0.28+1.00

id text identifier

This is a unique row identifier: all 52037 values are distinct (n_unique equals n) with zero nulls or duplicates. Values are 10-11 character single-token strings (len_min 10, len_max 11, one_word_rate 1.0, allcaps_rate 1.0), and the samples are numeric strings resembling 10-11 digit codes (e.g., FIPS-style geography IDs like 42069110300).

Treatment: Use as the primary key for joins; exclude from modelling features.

anthropic:claude-opus-4-7 · confidence high
Out[13]:

saturn.columns["id"].stats

statvalue
n52,037
nulls0 (0.0%)
unique52,037
len_min 10
len_max 11
len_mean 10.84
len_median 11
len_p95 11
word_mean 1
word_median 1
n_empty 0
n_duplicates 0
duplicate_rate 0
vocab_size 20,000
readability_flesch_mean 121.2
emoji_rate 0
url_rate 0
one_word_rate 1
allcaps_rate 1
boilerplate_rate 0
alert: near_unique100.0% of rows are unique strings
alert: one_word100.0% rows are a single word
alert: allcaps100.0% rows are all-caps
alert: short_text95th-percentile length under 20 chars
Fig 8.
Character-length distribution for id.
Show data table
charscount
10 – 108218
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 100
10 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 110
11 – 1143819

st categorical feature

This column holds US state names — 51 unique values (likely 50 states plus DC) across 52,037 rows with no nulls. The distribution roughly tracks population: Texas leads at 7.7%, followed by California, Florida, Ohio, and Pennsylvania. Entropy ratio of 0.915 indicates a fairly even spread with no single state dominating.

Treatment: One-hot or target-encode for modelling; consider grouping low-frequency states.

anthropic:claude-opus-4-7 · confidence high
Out[16]:

saturn.columns["st"].stats

statvalue
n52,037
nulls0 (0.0%)
unique51
top_value Texas
top_rate 0.07706
cardinality 51
entropy 5.192
entropy_ratio 0.9153
Fig 9.
Top values for st.
Show data table
valuecountshare
Texas40107.7%
California37277.2%
Florida30185.8%
Ohio23024.4%
Pennsylvania22424.3%
Michigan20734.0%
New York20594.0%
North Carolina19353.7%
Illinois19063.7%
Georgia17513.4%
Virginia14412.8%
Tennessee13582.6%
Indiana13282.6%
New Jersey12102.3%
Missouri11652.2%
Washington11272.2%
Wisconsin10922.1%
Minnesota10852.1%
Alabama10812.1%
Arizona10762.1%

cty text feature

This column ('cty') holds US county/parish names — short text averaging 2 words and 14 characters, with 'county' appearing in 19,399 rows and recurring names like Jefferson, Montgomery, and Maricopa County topping the list. With only 1,870 unique values across 52,037 rows, the duplicate rate is 96.4%, which is expected for a categorical geography field rather than a data-quality issue. No nulls, no URLs, no emoji — clean categorical text.

Treatment: Treat as a high-cardinality categorical; encode via target/frequency encoding or join to a county FIPS lookup.

anthropic:claude-opus-4-7 · confidence high
Out[19]:

saturn.columns["cty"].stats

statvalue
n52,037
nulls0 (0.0%)
unique1,870
len_min 10
len_max 33
len_mean 14.3
len_median 14
len_p95 18
word_mean 2.099
word_median 2
n_empty 0
n_duplicates 50,167
duplicate_rate 0.9641
vocab_size 1,651
readability_flesch_mean 25.91
emoji_rate 0
url_rate 0
one_word_rate 0
allcaps_rate 0
boilerplate_rate 0
alert: short_text95th-percentile length under 20 chars
alert: duplicates96.4% duplicate strings
Fig 10.
Character-length distribution for cty.
Show data table
charscount
10 – 11455
11 – 114694
11 – 120
12 – 125677
12 – 130
13 – 1310047
13 – 148997
14 – 150
15 – 157905
15 – 160
16 – 166217
16 – 170
17 – 174326
17 – 181685
18 – 190
19 – 191059
19 – 200
20 – 20169
20 – 210
21 – 22376
22 – 22254
22 – 230
23 – 2397
23 – 240
24 – 245
24 – 250
25 – 2634
26 – 266
26 – 270
27 – 2711
27 – 280
28 – 2818
28 – 290
29 – 300
30 – 300
30 – 310
31 – 312
31 – 320
32 – 320
32 – 333

pov numeric feature

Likely a poverty rate (percent) feature, bounded between 0 and 99.5 with a median of 10.8 and IQR 6.0-18.4. The distribution is right-skewed (skew 1.54, kurtosis 3.08) with 2116 outliers (4.07%) in the upper tail, and only 0.09% zeros across 693 distinct values over 52037 rows.

Treatment: Consider a log1p or winsorising transform before modelling to dampen the right-skewed tail.

anthropic:claude-opus-4-7 · confidence high
Out[22]:

saturn.columns["pov"].stats

statvalue
n52,037
nulls0 (0.0%)
unique693
min 0
max 99.5
mean 13.7
median 10.8
std 10.58
q1 6
q3 18.4
iqr 12.4
skew 1.536
kurtosis 3.077
n_outliers 2,116
outlier_rate 0.04066
zero_rate 0.0008648
Fig 11.
Distribution of pov. Vertical dash marks the median.
Show data table
bincount
0 – 2.4872933
2.487 – 4.9757061
4.975 – 7.4627249
7.462 – 9.956599
9.95 – 12.445536
12.44 – 14.924615
14.92 – 17.413752
17.41 – 19.92876
19.9 – 22.392480
22.39 – 24.881844
24.88 – 27.361480
27.36 – 29.851171
29.85 – 32.341029
32.34 – 34.82766
34.82 – 37.31607
37.31 – 39.8488
39.8 – 42.29399
42.29 – 44.77278
44.77 – 47.26222
47.26 – 49.75148
49.75 – 52.24114
52.24 – 54.72112
54.72 – 57.2178
57.21 – 59.748
59.7 – 62.1942
62.19 – 64.6736
64.67 – 67.1625
67.16 – 69.6514
69.65 – 72.1410
72.14 – 74.627
74.62 – 77.116
77.11 – 79.64
79.6 – 82.095
82.09 – 84.572
84.57 – 87.060
87.06 – 89.550
89.55 – 92.040
92.04 – 94.520
94.52 – 97.010
97.01 – 99.51

inc numeric feature

Likely an income or annual revenue figure: values range from 2,499 to 250,001 with a mean of 78,215 and median of 70,455. The distribution is right-skewed (skew 1.45) with about 3.9% outliers (2,047 rows), and the suspiciously round max of 250,001 hints at a censoring cap.

Treatment: log-transform and consider clipping at the 250,001 cap before modelling.

anthropic:claude-opus-4-7 · confidence high
Out[25]:

saturn.columns["inc"].stats

statvalue
n52,037
nulls0 (0.0%)
unique31,375
min 2,499
max 250,001
mean 7.821e+04
median 70,455
std 3.573e+04
q1 54,059
q3 94,375
iqr 40,316
skew 1.451
kurtosis 3.144
n_outliers 2,047
outlier_rate 0.03934
zero_rate 0
Fig 12.
Distribution of inc. Vertical dash marks the median.
Show data table
bincount
2499 – 86874
8687 – 1.487e+0438
1.487e+04 – 2.106e+04193
2.106e+04 – 2.725e+04634
2.725e+04 – 3.344e+041297
3.344e+04 – 3.962e+042050
3.962e+04 – 4.581e+043155
4.581e+04 – 5.2e+044048
5.2e+04 – 5.819e+044797
5.819e+04 – 6.437e+045037
6.437e+04 – 7.056e+044829
7.056e+04 – 7.675e+044317
7.675e+04 – 8.294e+043593
8.294e+04 – 8.912e+042932
8.912e+04 – 9.531e+042448
9.531e+04 – 1.015e+052120
1.015e+05 – 1.077e+051859
1.077e+05 – 1.139e+051463
1.139e+05 – 1.201e+051241
1.201e+05 – 1.262e+051079
1.262e+05 – 1.324e+05910
1.324e+05 – 1.386e+05701
1.386e+05 – 1.448e+05506
1.448e+05 – 1.51e+05484
1.51e+05 – 1.572e+05382
1.572e+05 – 1.634e+05361
1.634e+05 – 1.696e+05261
1.696e+05 – 1.758e+05178
1.758e+05 – 1.819e+05171
1.819e+05 – 1.881e+05131
1.881e+05 – 1.943e+05141
1.943e+05 – 2.005e+05106
2.005e+05 – 2.067e+05100
2.067e+05 – 2.129e+0563
2.129e+05 – 2.191e+0564
2.191e+05 – 2.253e+0562
2.253e+05 – 2.314e+0540
2.314e+05 – 2.376e+0529
2.376e+05 – 2.438e+0532
2.438e+05 – 2.5e+05181

dist_share numeric feature

A numeric share/proportion field bounded between 0 and 10000, suggesting basis-point or per-myriad encoding rather than a 0-1 ratio. The distribution is nearly symmetric (skew -0.09) but distinctly flat with negative kurtosis (-1.51) and an IQR spanning 1785.6 to 9377.5, pointing to a near-uniform spread across the range rather than a peaked distribution. About 2.1% of rows are exactly zero, and there are no statistical outliers.

Treatment: Rescale to [0,1] by dividing by 10000 before modelling; consider a separate zero-flag for the 2% exact zeros.

anthropic:claude-opus-4-7 · confidence high
Out[28]:

saturn.columns["dist_share"].stats

statvalue
n52,037
nulls0 (0.0%)
unique33,000
min 0
max 10,000
mean 5396
median 5503
std 3671
q1 1786
q3 9378
iqr 7592
skew -0.09003
kurtosis -1.506
n_outliers 0
outlier_rate 0
zero_rate 0.02056
Fig 13.
Distribution of dist_share. Vertical dash marks the median.
Show data table
bincount
0 – 2505132
250 – 5001829
500 – 7501421
750 – 10001277
1000 – 12501119
1250 – 15001073
1500 – 17501017
1750 – 2000973
2000 – 2250976
2250 – 2500907
2500 – 2750894
2750 – 3000893
3000 – 3250857
3250 – 3500862
3500 – 3750840
3750 – 4000865
4000 – 4250822
4250 – 4500829
4500 – 4750876
4750 – 5000844
5000 – 5250871
5250 – 5500834
5500 – 5750839
5750 – 6000845
6000 – 6250767
6250 – 6500798
6500 – 6750779
6750 – 7000878
7000 – 7250872
7250 – 7500803
7500 – 7750788
7750 – 8000843
8000 – 8250810
8250 – 8500861
8500 – 8750787
8750 – 9000871
9000 – 9250952
9250 – 95001070
9500 – 97501401
9750 – 1e+0411062

pop numeric feature

This is a numeric 'pop' column with 52037 non-null values, likely a population count (or similar headcount metric) per row, ranging from 102 to 37452 with a median of 4169. The distribution is right-skewed (skew 1.68, kurtosis 9.99) with 972 outliers (1.87%) above the upper tail, and there are no zeros or nulls. With 8732 unique values across 52037 rows, the same population figures repeat frequently, suggesting the same entity appears across multiple rows.

Treatment: Log-transform before modelling to dampen the right skew and outliers.

anthropic:claude-opus-4-7 · confidence medium
Out[31]:

saturn.columns["pop"].stats

statvalue
n52,037
nulls0 (0.0%)
unique8,732
min 102
max 37,452
mean 4432
median 4,169
std 2025
q1 3,022
q3 5,523
iqr 2,501
skew 1.684
kurtosis 9.986
n_outliers 972
outlier_rate 0.01868
zero_rate 0
Fig 14.
Distribution of pop. Vertical dash marks the median.
Show data table
bincount
102 – 1036355
1036 – 19703438
1970 – 29037964
2903 – 383710408
3837 – 477110267
4771 – 57047885
5704 – 66385418
6638 – 75723139
7572 – 85061585
8506 – 9440706
9440 – 1.037e+04326
1.037e+04 – 1.131e+04164
1.131e+04 – 1.224e+04127
1.224e+04 – 1.317e+0466
1.317e+04 – 1.411e+0444
1.411e+04 – 1.504e+0432
1.504e+04 – 1.598e+0437
1.598e+04 – 1.691e+0421
1.691e+04 – 1.784e+0414
1.784e+04 – 1.878e+044
1.878e+04 – 1.971e+047
1.971e+04 – 2.064e+048
2.064e+04 – 2.158e+043
2.158e+04 – 2.251e+044
2.251e+04 – 2.345e+043
2.345e+04 – 2.438e+041
2.438e+04 – 2.531e+042
2.531e+04 – 2.625e+041
2.625e+04 – 2.718e+041
2.718e+04 – 2.811e+040
2.811e+04 – 2.905e+043
2.905e+04 – 2.998e+041
2.998e+04 – 3.092e+040
3.092e+04 – 3.185e+040
3.185e+04 – 3.278e+040
3.278e+04 – 3.372e+042
3.372e+04 – 3.465e+040
3.465e+04 – 3.558e+040
3.558e+04 – 3.652e+040
3.652e+04 – 3.745e+041

snap numeric feature

A right-skewed numeric feature with values spanning 0 to 1888 and a median of 146, well below the mean of 191. Skew of 1.71 and kurtosis of 4.70 confirm a long upper tail, with 1918 outliers (3.69%) and 2.28% zeros. The 1034 unique values across 52037 rows suggest a bounded count or score rather than a continuous measurement.

Treatment: log-transform or winsorize before regression to tame the right tail.

anthropic:claude-opus-4-7 · confidence medium
Out[34]:

saturn.columns["snap"].stats

statvalue
n52,037
nulls0 (0.0%)
unique1,034
min 0
max 1,888
mean 191
median 146
std 170.2
q1 66
q3 268
iqr 202
skew 1.709
kurtosis 4.702
n_outliers 1,918
outlier_rate 0.03686
zero_rate 0.02279
Fig 15.
Distribution of snap. Vertical dash marks the median.
Show data table
bincount
0 – 47.29587
47.2 – 94.48464
94.4 – 141.67361
141.6 – 188.86046
188.8 – 2364810
236 – 283.23869
283.2 – 330.42938
330.4 – 377.62328
377.6 – 424.81718
424.8 – 4721286
472 – 519.2980
519.2 – 566.4673
566.4 – 613.6492
613.6 – 660.8417
660.8 – 708288
708 – 755.2213
755.2 – 802.4139
802.4 – 849.6119
849.6 – 896.877
896.8 – 94468
944 – 991.245
991.2 – 103829
1038 – 108622
1086 – 113319
1133 – 11809
1180 – 122710
1227 – 12746
1274 – 13225
1322 – 13697
1369 – 14160
1416 – 14631
1463 – 15102
1510 – 15581
1558 – 16050
1605 – 16521
1652 – 16991
1699 – 17462
1746 – 17942
1794 – 18411
1841 – 18881

How to cite

click to copy

BibTeX
@misc{saturn-environmental-desert-data-2026,
  author       = {Steuber, Luke},
  title        = {Saturn reading: environmental desert data},
  year         ={2026},
  howpublished = {\url{https://dr.eamer.dev/saturn/view/environmental-desert_data}},
  note         = {Profiled with saturn-dissect v0.2.0, prompt saturn-insight-v2, model anthropic:claude-opus-4-7},
}
APA
Steuber, L. (2026). Saturn reading: environmental desert data. Source: /home/coolhand/html/datavis/data_trove/data/environmental/desert_data.json. Profiled with saturn-dissect v0.2.0 (saturn-insight-v2, anthropic:claude-opus-4-7). Retrieved from https://dr.eamer.dev/saturn/view/environmental-desert_data