saturn

This dataset contains 2022 US Census disability counts for 3,222 counties, broken out by disability type (ambulatory, cognitive, hearing, vision, self-care, independent living) along with totals, a derived disability rate, and FIPS identifiers. Nearly every count column is heavily right-skewed (skew above 10) with substantial outliers — total_population alone ranges from 47 to 9.87M with a mean of ~102K but a median of just 25,328, so a handful of large counties dominate the raw counts. The disability_rate field is the most analyst-friendly view: it's bounded, less skewed (skew 2.17), and centers around a median of 1.07 with an IQR of 0.77–1.42. Start with disability_rate to compare counties on equal footing, then look at total_population to understand the size distribution before interpreting any raw disability counts.

This column is the FIPS county code: every one of the 3222 rows is unique and non-null, and the value range (1001 to 72153) matches the standard US state+county FIPS encoding. The distribution is near-symmetric (skew 0.16, kurtosis -0.63) with no outliers, which is expected for an identifier rather than a measured quantity. Treat it as a categorical key, not a number.

Each of the 3,222 rows holds a unique county-and-state string (e.g., '... County, Texas'), averaging 24 characters and roughly 3 words. The token 'county,' appears 2,999 times, so a small minority of entries use a different suffix (parish, borough, census area). Texas (256), Virginia (189), and Georgia (159) lead the state distribution, consistent with a full U.S. county roster.

This is almost certainly the US state FIPS code: 52 unique integer values across 3222 rows, ranging from 1 to 72 with no nulls or zeros. The count of 52 (rather than 50) and a max of 72 indicate inclusion of DC and territories like Puerto Rico. Distribution is roughly uniform (skew 0.16, kurtosis -0.63), consistent with a categorical geographic identifier rather than a measurement.

This is the county-level portion of a FIPS code, stored as an integer from 1 to 840 across 3222 rows with no nulls and only 330 distinct values. The distribution is heavily right-skewed (skew 2.87, kurtosis 11.64) with 178 outliers (5.5%), which is expected since county codes are categorical identifiers and most values cluster low (median 79, Q3 133) while a few counties carry much larger codes. Treating this as a numeric feature would be misleading despite the numeric dtype.

Likely a county- or region-level total population count across 3,222 rows with no nulls and 3,141 unique values. The distribution is extremely right-skewed (skew 13.38, kurtosis 298.69): the median is 25,328 but the mean is 102,232 and the max reaches 9,866,623, with 453 outliers (14.06%) flagged above the IQR fence.

A heavily right-skewed count of disability cases or claims, ranging from 0 to 69,705 with a median of just 298 but a mean of 1,043. Skew of 10.28 and kurtosis of 166.8 indicate an extreme long tail, with 404 outliers (12.5% of rows) and only 1.7% zeros. The std (2,906) dwarfs the IQR (689), so a small number of very large records dominate the distribution.

Numeric disability_rate spanning 0.0 to 9.17 with a median of 1.07 and IQR 0.77-1.42, almost certainly a per-row rate or percentage. The distribution is heavily right-skewed (skew 2.17, kurtosis 15.24) with 117 outliers (3.6%) stretching well beyond the typical range, and 1.7% of rows sit at exactly zero. No nulls across 3,222 rows, and only 305 distinct values suggest rounding to two decimals.

Counts of people recorded as having no disability per geographic or administrative unit, ranging from 0 to 2,091,332 with a median of 5,607. The distribution is extremely right-skewed (skew 12.67, kurtosis 259.77) and the mean of 22,872 sits well above Q3 of 14,739, with 442 outliers (13.7%) flagging a long tail of very large units. Only one zero is present and there are no nulls, so the heavy tail—not missingness—is the dominant feature.

This column appears to be a count of people with one disability per geographic or administrative unit, ranging from 0 to 44,466 with a median of 217.5. The distribution is severely right-skewed (skew 9.45, kurtosis 139.4), with the mean (755.7) more than triple the median and 408 outliers (12.7% of rows) — consistent with a few very large units dominating a long tail of small ones. About 2.8% of rows are zero and there are no nulls.

This column appears to be a count of people (likely per geographic unit) reporting two or more disabilities, ranging from 0 to 25,239 with a median of just 76. The distribution is extremely right-skewed (skew 12.57, kurtosis 253.95), with 11.67% of rows flagged as outliers and ~9% exact zeros, suggesting a few very large jurisdictions dominate while most are small. The mean (287.7) sits well above Q3 (222), confirming the long tail.

This appears to be a count or population-style measure related to hearing disability, with all 3222 rows populated and 2314 distinct values ranging from 1 to 296898. The distribution is extremely right-skewed (skew 11.54, kurtosis 226.6) with a median of 1326 well below the mean of 4003, and 391 outliers (12.1%) inflate the tail. The min of 1 and absence of zeros suggest these are aggregated counts rather than individual indicators.

Numeric counts of people with a vision disability per geographic or demographic unit, ranging from 0 to 346,901 with a median of 1,361. The distribution is extremely right-skewed (skew 12.29, kurtosis 254.79) and the mean of 4,246 sits well above Q3 of 3,291, with 380 outliers (11.8%) inflating the upper tail. Near-zero zero_rate (0.03%) and no nulls suggest clean population-style aggregates rather than survey responses.

Likely a count of people with a cognitive disability per geographic or administrative unit, ranging from 0 to 413,990 with a median of 1,623. The distribution is severely right-skewed (skew 12.09, kurtosis 254.7) with 375 outliers (11.6% of rows) and a mean (5,142) more than triple the median, indicating a few very large units dominate. Near-zero null and zero rates suggest the count is reliably populated.

Counts of people with ambulatory disability per geographic unit, ranging from 3 to 548,175 with a median of 2,197. The distribution is severely right-skewed (skew 13.0, kurtosis 288.7) and 11.4% of rows are flagged as outliers, indicating a long tail of very large jurisdictions dominating the mean (6,497) versus the median.

Numeric counts of people with a self-care disability, likely aggregated per geographic or demographic unit given the 3222 rows and 1961 unique values. The distribution is severely right-skewed (skew 16.8, kurtosis 478.7) with a median of 772.5 but a max of 281,611, and 355 outliers (11.0% of rows) sit far above the Q3 of 1948.5. Near-zero null and zero rates suggest the field is consistently populated.

Counts of people with an independent-living disability per geographic unit, ranging from 2 to 1,417,825 with a median of 3,135. The distribution is severely right-skewed (skew 14.09, kurtosis 329.97) and 13.8% of rows (445) flag as outliers, suggesting the column mixes small areas with very large aggregates. No nulls or zeros are present.