saturn·

← all readings

data trove onion headlines

source /home/coolhand/html/datavis/data_trove/entertainment/satire/theonion_index_to_dataset.csv 2,103 rows 3 columns profiled 2026-06-21 raw JSON static .html .ipynb Report Notebook

Reading

dataset summary · medium confidence anthropic:default

This dataset is an index of 2,103 articles from The Onion, a satirical news outlet, containing article headlines, thumbnail image URLs, and a sequential row ID. The headline column is the most analytically interesting field, with a vocabulary of 7,613 unique words and mean headline length of about 9 words and 61 characters — worth exploring for length distribution and common vocabulary patterns. There are also 209 duplicate image URLs (~10% of rows), suggesting some thumbnails are reused across multiple articles, with one image appearing 11 times.

citing: row_count · column_count · n_unique · duplicate_rate · n_duplicates · vocab_size · len_mean · len_median · len_max · word_mean

https://i.kinja-img.com/gawker-media/image/upload/c_fit,f_auto,g_center,q_60,w_645/a30d02609da7e7ddb46d6edea9460d5e.jpg · Most-reused thumbnail URLs show one image appears 11 times — check whether repeated images indicate article templates or content categories.
Show data table
charscount
107 – 1071948
107 – 1080
108 – 1080
108 – 1080
108 – 1080
108 – 1090
109 – 1090
109 – 1090
109 – 1100
110 – 1100
110 – 1100
110 – 1110
111 – 1110
111 – 1110
111 – 1120
112 – 1120
112 – 1120
112 – 1120
112 – 1130
113 – 1130
113 – 1130
113 – 1140
114 – 1140
114 – 1140
114 – 1140
114 – 1150
115 – 1150
115 – 1150
115 – 1160
116 – 1160
116 – 1160
116 – 1170
117 – 1170
117 – 1170
117 – 1180
118 – 1180
118 – 1180
118 – 1180
118 – 1190
119 – 119155
1 · The row ID is perfectly uniform with no gaps or outliers, confirming this is a clean sequential index with no missing records.
Show data table
bincount
2 – 54.5553
54.55 – 107.153
107.1 – 159.652
159.6 – 212.253
212.2 – 264.852
264.8 – 317.353
317.3 – 369.852
369.8 – 422.453
422.4 – 474.952
474.9 – 527.553
527.5 – 58053
580 – 632.652
632.6 – 685.153
685.1 – 737.752
737.7 – 790.253
790.2 – 842.852
842.8 – 895.353
895.3 – 947.952
947.9 – 100053
1000 – 105352
1053 – 110653
1106 – 115853
1158 – 121152
1211 – 126353
1263 – 131652
1316 – 136853
1368 – 142152
1421 – 147353
1473 – 152652
1526 – 157853
1578 – 163153
1631 – 168452
1684 – 173653
1736 – 178952
1789 – 184153
1841 – 189452
1894 – 194653
1946 – 199952
1999 – 205153
2051 – 210453

Schema

3 columns
Alerts
1 numeric 0.0% 2,103
‘That’ll Be $3,’ Says Trump After Handing Water Bottle To Sick Ohio Resident text 0.0% 2,095
near_unique
https://i.kinja-img.com/gawker-media/image/upload/c_fit,f_auto,g_center,q_60,w_645/a30d02609da7e7ddb46d6edea9460d5e.jpg text 0.0% 1,894
one_word url_heavy

1

numeric identifier
This column is almost certainly a row index or sequential identifier: it contains 2103 values, all unique, with min=2 and max=2104, perfectly symmetric distribution (skew=0.0, mean=median=1053.0), and zero outliers. The near-perfect uniformity (platykurtic at -1.2, IQR=1051 spanning exactly half the range) confirms an arithmetic sequence. The column name '1' suggests it was emitted as an unnamed index column during export. Treatment: Drop before modelling; carries no predictive signal and is a row index artifact. high · anthropic:default
n
2,103
nulls
0 (0.0%)
unique
2,103
min
2
max
2,104
mean
1,053
median
1,053
std
607.2
q1
527.5
q3
1578
iqr
1,051
skew
0
kurtosis
-1.2
n_outliers
0
outlier_rate
0
zero_rate
0

‘That’ll Be $3,’ Says Trump After Handing Water Bottle To Sick Ohio Resident

text free_text near_unique
This column contains satirical or parody news headlines, with the column name itself being a representative example ('That'll Be $3,' Says Trump After Handing Water Bottle To Sick Ohio Resident). With 2095 unique values out of 2103 rows and a mean length of ~61 characters (~9 words), these are near-unique short text strings consistent with headline-style content. Surprisingly, 8 duplicate headlines exist despite the near-unique alert, and the Flesch readability score of 46.87 indicates moderately difficult prose — higher than typical clickbait but consistent with crafted satirical writing. Top words ('to', 'of', 'in', 'new') are generic function words offering little discriminative signal on their own. Treatment: Tokenize and embed (e.g., sentence-transformers) before modelling; deduplicate the 8 duplicate headlines first. high · anthropic:default
n
2,103
nulls
0 (0.0%)
unique
2,095
len_min
8
len_max
926
len_mean
60.99
len_median
58
len_p95
102
word_mean
9.268
word_median
9
n_empty
0
n_duplicates
8
duplicate_rate
0.003804
vocab_size
7,613
readability_flesch_mean
46.87
emoji_rate
0
url_rate
0
one_word_rate
0.0004755
allcaps_rate
0.0004755
boilerplate_rate
0.0004755

https://i.kinja-img.com/gawker-media/image/upload/c_fit,f_auto,g_center,q_60,w_645/a30d02609da7e7ddb46d6edea9460d5e.jpg

text metadata one_word url_heavy
This column contains Kinja/Gawker CDN image URLs, serving as thumbnail or article image references across 2103 rows. Every value is a URL (url_rate 1.0) pointing to a fixed-size JPEG (w_645, q_60 transform parameters are baked into all URLs), making the path prefix structurally identical across all rows with only the hash filename varying. Surprising: 209 duplicate URLs exist (duplicate_rate ~9.9%), with one image URL appearing 11 times, suggesting multiple articles or rows share the same thumbnail — likely a placeholder or default image reuse pattern. Treatment: Extract the hash filename as a unique image identifier; flag the top duplicate URL (count=11) as a likely default/placeholder and treat separately from article-specific images. high · anthropic:default
n
2,103
nulls
0 (0.0%)
unique
1,894
len_min
107
len_max
119
len_mean
107.9
len_median
107
len_p95
119
word_mean
1
word_median
1
n_empty
0
n_duplicates
209
duplicate_rate
0.09938
vocab_size
1,894
readability_flesch_mean
-1672
emoji_rate
0
url_rate
1
one_word_rate
1
allcaps_rate
0
boilerplate_rate
0