ml 32m links

saturn notebook · generated 2026-05-01 Report Notebook

Overview

Source: /home/coolhand/html/datavis/data_trove/entertainment/movies/ml-32m/links.csv

Saturn profiled 87,585 rows across 3 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/entertainment/movies/ml-32m/links.csv",
    "--findings", "ml-32m-links.json",
    "--llm", "anthropic:claude-opus-4-7",
])

Summary confidence: high

This is a movie ID linkage table with 87,585 rows and 3 numeric identifier columns: imdbId, tmdbId, and movieId. As expected for ID columns, movieId and imdbId are unique per row, while tmdbId is nearly unique with a tiny null rate (0.0014). The most notable shape is imdbId, which is heavily right-skewed (skew 2.24) with about 8.6% flagged as outliers — reflecting how IMDb IDs span a huge range from 1 up to ~29M. tmdbId is also right-skewed but more moderately, while movieId is distributed fairly evenly up to 292,757. There's little analytical signal here beyond confirming the file is a clean ID crosswalk.

citing: row_count · column_count · columns.imdbId.stats.skew · columns.imdbId.stats.outlier_rate · columns.imdbId.stats.max · columns.tmdbId.stats.skew · columns.tmdbId.null_rate · columns.movieId.stats.max · columns.movieId.stats.skew

Out[4]:

saturn.schema() · 3 columns

column	kind	n	null%	unique	alerts
movieId	numeric	87,585	0.0%	87,585
imdbId	numeric	87,585	0.0%	87,585	high_skew outliers
tmdbId	numeric	87,585	0.1%	87,425

Fig 1.

imdbId · Heavy right skew and a long tail up to ~29M — most IDs cluster well below the mean.

Show data table

Histogram bins for imdbId (median: 492996.0).
bin	count
1 – 7.27e+05	44057
7.27e+05 – 1.454e+06	7224
1.454e+06 – 2.181e+06	6540
2.181e+06 – 2.908e+06	3749
2.908e+06 – 3.635e+06	3128
3.635e+06 – 4.362e+06	2659
4.362e+06 – 5.089e+06	2286
5.089e+06 – 5.816e+06	2278
5.816e+06 – 6.543e+06	2216
6.543e+06 – 7.27e+06	1712
7.27e+06 – 7.997e+06	1536
7.997e+06 – 8.724e+06	1290
8.724e+06 – 9.451e+06	1232
9.451e+06 – 1.018e+07	839
1.018e+07 – 1.091e+07	1095
1.091e+07 – 1.163e+07	930
1.163e+07 – 1.236e+07	678
1.236e+07 – 1.309e+07	601
1.309e+07 – 1.381e+07	674
1.381e+07 – 1.454e+07	572
1.454e+07 – 1.527e+07	531
1.527e+07 – 1.599e+07	391
1.599e+07 – 1.672e+07	156
1.672e+07 – 1.745e+07	96
1.745e+07 – 1.818e+07	79
1.818e+07 – 1.89e+07	83
1.89e+07 – 1.963e+07	56
1.963e+07 – 2.036e+07	120
2.036e+07 – 2.108e+07	85
2.108e+07 – 2.181e+07	173
2.181e+07 – 2.254e+07	135
2.254e+07 – 2.326e+07	58
2.326e+07 – 2.399e+07	55
2.399e+07 – 2.472e+07	36
2.472e+07 – 2.545e+07	18
2.545e+07 – 2.617e+07	15
2.617e+07 – 2.69e+07	65
2.69e+07 – 2.763e+07	59
2.763e+07 – 2.835e+07	52
2.835e+07 – 2.908e+07	26

Fig 2.

tmdbId · Moderately right-skewed distribution of TMDB IDs, with values concentrated under 400K.

Show data table

Histogram bins for tmdbId (median: 139272.0).
bin	count
2 – 2.966e+04	13785
2.966e+04 – 5.932e+04	12892
5.932e+04 – 8.898e+04	8691
8.898e+04 – 1.186e+05	5333
1.186e+05 – 1.483e+05	4232
1.483e+05 – 1.78e+05	3219
1.78e+05 – 2.076e+05	2874
2.076e+05 – 2.373e+05	2446
2.373e+05 – 2.669e+05	2936
2.669e+05 – 2.966e+05	2854
2.966e+05 – 3.262e+05	2029
3.262e+05 – 3.559e+05	2342
3.559e+05 – 3.856e+05	2233
3.856e+05 – 4.152e+05	2131
4.152e+05 – 4.449e+05	2112
4.449e+05 – 4.745e+05	2038
4.745e+05 – 5.042e+05	1694
5.042e+05 – 5.339e+05	1579
5.339e+05 – 5.635e+05	1258
5.635e+05 – 5.932e+05	1452
5.932e+05 – 6.228e+05	1212
6.228e+05 – 6.525e+05	921
6.525e+05 – 6.821e+05	939
6.821e+05 – 7.118e+05	534
7.118e+05 – 7.415e+05	715
7.415e+05 – 7.711e+05	670
7.711e+05 – 8.008e+05	684
8.008e+05 – 8.304e+05	529
8.304e+05 – 8.601e+05	532
8.601e+05 – 8.898e+05	483
8.898e+05 – 9.194e+05	340
9.194e+05 – 9.491e+05	404
9.491e+05 – 9.787e+05	322
9.787e+05 – 1.008e+06	287
1.008e+06 – 1.038e+06	262
1.038e+06 – 1.068e+06	166
1.068e+06 – 1.097e+06	129
1.097e+06 – 1.127e+06	83
1.127e+06 – 1.157e+06	82
1.157e+06 – 1.186e+06	37

Fig 3.

movieId · MovieLens IDs spread fairly evenly across the range, slightly weighted toward higher values.

Show data table

Histogram bins for movieId (median: 165741.0).
bin	count
1 – 7320	7196
7320 – 1.464e+04	1111
1.464e+04 – 2.196e+04	0
2.196e+04 – 2.928e+04	1114
2.928e+04 – 3.66e+04	796
3.66e+04 – 4.391e+04	435
4.391e+04 – 5.123e+04	754
5.123e+04 – 5.855e+04	816
5.855e+04 – 6.587e+04	789
6.587e+04 – 7.319e+04	1132
7.319e+04 – 8.051e+04	1107
8.051e+04 – 8.783e+04	1398
8.783e+04 – 9.515e+04	1556
9.515e+04 – 1.025e+05	1541
1.025e+05 – 1.098e+05	1535
1.098e+05 – 1.171e+05	1847
1.171e+05 – 1.244e+05	2673
1.244e+05 – 1.317e+05	2763
1.317e+05 – 1.391e+05	3280
1.391e+05 – 1.464e+05	3282
1.464e+05 – 1.537e+05	3146
1.537e+05 – 1.61e+05	3258
1.61e+05 – 1.683e+05	3464
1.683e+05 – 1.757e+05	3492
1.757e+05 – 1.83e+05	3484
1.83e+05 – 1.903e+05	3486
1.903e+05 – 1.976e+05	3401
1.976e+05 – 2.049e+05	3369
2.049e+05 – 2.122e+05	3050
2.122e+05 – 2.196e+05	2907
2.196e+05 – 2.269e+05	2325
2.269e+05 – 2.342e+05	1839
2.342e+05 – 2.415e+05	1672
2.415e+05 – 2.488e+05	1531
2.488e+05 – 2.562e+05	1566
2.562e+05 – 2.635e+05	1679
2.635e+05 – 2.708e+05	1911
2.708e+05 – 2.781e+05	2168
2.781e+05 – 2.854e+05	2580
2.854e+05 – 2.928e+05	2132

Fig 4.

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 %
movieId	numeric	0.0%
imdbId	numeric	0.0%
tmdbId	numeric	0.1%

Fig 5.

Pearson correlation across numeric columns (sampled, bounded).

Show data table

Pearson correlation across 3 numeric columns (values clipped to 2 decimals).
	movieId	imdbId	tmdbId
movieId	+1.00	+0.51	+0.64
imdbId	+0.51	+1.00	+0.44
tmdbId	+0.64	+0.44	+1.00

movieId numeric identifier

movieId is a fully unique integer key spanning 1 to 292,757 across all 87,585 rows with no nulls or zeros, consistent with an external catalogue identifier rather than a dense surrogate index. The range far exceeds the row count, indicating gaps in the ID space (likely a sparse catalogue like MovieLens). Distribution stats (mean 157,651, median 165,741, skew -0.39) are artefacts of the ID assignment and carry no analytical meaning.

Treatment: Use as a join key to movie metadata; exclude from modelling features.

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

Out[11]:

saturn.columns["movieId"].stats

stat	value
n	87,585
nulls	0 (0.0%)
unique	87,585
min	1
max	292,757
mean	1.577e+05
median	165,741
std	7.901e+04
q1	112,657
q3	213,203
iqr	100,546
skew	-0.3914
kurtosis	-0.5775
n_outliers	0
outlier_rate	0
zero_rate	0

Fig 6.

Distribution of movieId. Vertical dash marks the median.

Show data table

Histogram bins for movieId (median: 165741.0).
bin	count
1 – 7320	7196
7320 – 1.464e+04	1111
1.464e+04 – 2.196e+04	0
2.196e+04 – 2.928e+04	1114
2.928e+04 – 3.66e+04	796
3.66e+04 – 4.391e+04	435
4.391e+04 – 5.123e+04	754
5.123e+04 – 5.855e+04	816
5.855e+04 – 6.587e+04	789
6.587e+04 – 7.319e+04	1132
7.319e+04 – 8.051e+04	1107
8.051e+04 – 8.783e+04	1398
8.783e+04 – 9.515e+04	1556
9.515e+04 – 1.025e+05	1541
1.025e+05 – 1.098e+05	1535
1.098e+05 – 1.171e+05	1847
1.171e+05 – 1.244e+05	2673
1.244e+05 – 1.317e+05	2763
1.317e+05 – 1.391e+05	3280
1.391e+05 – 1.464e+05	3282
1.464e+05 – 1.537e+05	3146
1.537e+05 – 1.61e+05	3258
1.61e+05 – 1.683e+05	3464
1.683e+05 – 1.757e+05	3492
1.757e+05 – 1.83e+05	3484
1.83e+05 – 1.903e+05	3486
1.903e+05 – 1.976e+05	3401
1.976e+05 – 2.049e+05	3369
2.049e+05 – 2.122e+05	3050
2.122e+05 – 2.196e+05	2907
2.196e+05 – 2.269e+05	2325
2.269e+05 – 2.342e+05	1839
2.342e+05 – 2.415e+05	1672
2.415e+05 – 2.488e+05	1531
2.488e+05 – 2.562e+05	1566
2.562e+05 – 2.635e+05	1679
2.635e+05 – 2.708e+05	1911
2.708e+05 – 2.781e+05	2168
2.781e+05 – 2.854e+05	2580
2.854e+05 – 2.928e+05	2132

imdbId numeric identifier

This is the IMDb title identifier: all 87,585 rows are unique with no nulls, matching the row count exactly. Although flagged as skewed (2.24) with 7,575 outliers, that's expected since IMDb IDs are issued sequentially over time and span 1 to 29,081,098. The distribution shape is an artifact of ID assignment, not a meaningful numeric signal.

Treatment: Treat as a key for joins to IMDb metadata; do not use as a numeric feature.

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

Out[14]:

saturn.columns["imdbId"].stats

stat	value
n	87,585
nulls	0 (0.0%)
unique	87,585
min	1
max	2.908e+07
mean	2.793e+06
median	492,996
std	4.279e+06
q1	94,642
q3	3.877e+06
iqr	3.783e+06
skew	2.243
kurtosis	5.857
n_outliers	7,575
outlier_rate	0.08649
zero_rate	0
alert: high_skew	skew=+2.24
alert: outliers	8.6% rows beyond 1.5 IQR

Fig 7.

Distribution of imdbId. Vertical dash marks the median.

Show data table

Histogram bins for imdbId (median: 492996.0).
bin	count
1 – 7.27e+05	44057
7.27e+05 – 1.454e+06	7224
1.454e+06 – 2.181e+06	6540
2.181e+06 – 2.908e+06	3749
2.908e+06 – 3.635e+06	3128
3.635e+06 – 4.362e+06	2659
4.362e+06 – 5.089e+06	2286
5.089e+06 – 5.816e+06	2278
5.816e+06 – 6.543e+06	2216
6.543e+06 – 7.27e+06	1712
7.27e+06 – 7.997e+06	1536
7.997e+06 – 8.724e+06	1290
8.724e+06 – 9.451e+06	1232
9.451e+06 – 1.018e+07	839
1.018e+07 – 1.091e+07	1095
1.091e+07 – 1.163e+07	930
1.163e+07 – 1.236e+07	678
1.236e+07 – 1.309e+07	601
1.309e+07 – 1.381e+07	674
1.381e+07 – 1.454e+07	572
1.454e+07 – 1.527e+07	531
1.527e+07 – 1.599e+07	391
1.599e+07 – 1.672e+07	156
1.672e+07 – 1.745e+07	96
1.745e+07 – 1.818e+07	79
1.818e+07 – 1.89e+07	83
1.89e+07 – 1.963e+07	56
1.963e+07 – 2.036e+07	120
2.036e+07 – 2.108e+07	85
2.108e+07 – 2.181e+07	173
2.181e+07 – 2.254e+07	135
2.254e+07 – 2.326e+07	58
2.326e+07 – 2.399e+07	55
2.399e+07 – 2.472e+07	36
2.472e+07 – 2.545e+07	18
2.545e+07 – 2.617e+07	15
2.617e+07 – 2.69e+07	65
2.69e+07 – 2.763e+07	59
2.763e+07 – 2.835e+07	52
2.835e+07 – 2.908e+07	26

tmdbId numeric foreign_key

This is almost certainly the TMDB movie/title identifier used for joins to The Movie Database. With 87,425 unique values across 87,585 rows and a null rate of 0.0014, it behaves as a near-unique key rather than a modelling feature, despite its numeric stats showing a wide range (2 to 1,186,337) and right skew (1.25).

Treatment: left-join on this id; do not use as a numeric feature.

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

Out[17]:

saturn.columns["tmdbId"].stats

stat	value
n	87,585
nulls	124 (0.1%)
unique	87,425
min	2
max	1.186e+06
mean	2.414e+05
median	139,272
std	2.471e+05
q1	46,836
q3	381,693
iqr	334,857
skew	1.254
kurtosis	0.9079
n_outliers	2,200
outlier_rate	0.02515
zero_rate	0

Fig 8.

Distribution of tmdbId. Vertical dash marks the median.

Show data table

Histogram bins for tmdbId (median: 139272.0).
bin	count
2 – 2.966e+04	13785
2.966e+04 – 5.932e+04	12892
5.932e+04 – 8.898e+04	8691
8.898e+04 – 1.186e+05	5333
1.186e+05 – 1.483e+05	4232
1.483e+05 – 1.78e+05	3219
1.78e+05 – 2.076e+05	2874
2.076e+05 – 2.373e+05	2446
2.373e+05 – 2.669e+05	2936
2.669e+05 – 2.966e+05	2854
2.966e+05 – 3.262e+05	2029
3.262e+05 – 3.559e+05	2342
3.559e+05 – 3.856e+05	2233
3.856e+05 – 4.152e+05	2131
4.152e+05 – 4.449e+05	2112
4.449e+05 – 4.745e+05	2038
4.745e+05 – 5.042e+05	1694
5.042e+05 – 5.339e+05	1579
5.339e+05 – 5.635e+05	1258
5.635e+05 – 5.932e+05	1452
5.932e+05 – 6.228e+05	1212
6.228e+05 – 6.525e+05	921
6.525e+05 – 6.821e+05	939
6.821e+05 – 7.118e+05	534
7.118e+05 – 7.415e+05	715
7.415e+05 – 7.711e+05	670
7.711e+05 – 8.008e+05	684
8.008e+05 – 8.304e+05	529
8.304e+05 – 8.601e+05	532
8.601e+05 – 8.898e+05	483
8.898e+05 – 9.194e+05	340
9.194e+05 – 9.491e+05	404
9.491e+05 – 9.787e+05	322
9.787e+05 – 1.008e+06	287
1.008e+06 – 1.038e+06	262
1.038e+06 – 1.068e+06	166
1.068e+06 – 1.097e+06	129
1.097e+06 – 1.127e+06	83
1.127e+06 – 1.157e+06	82
1.157e+06 – 1.186e+06	37

How to cite

click to copy

BibTeX

@misc{saturn-ml-32m-links-2026,
  author       = {Steuber, Luke},
  title        = {Saturn reading: ml 32m links},
  year         ={2026},
  howpublished = {\url{https://dr.eamer.dev/saturn/view/ml-32m-links}},
  note         = {Profiled with saturn-dissect v0.2.0, prompt saturn-insight-v2, model anthropic:claude-opus-4-7},
}

APA

Steuber, L. (2026). Saturn reading: ml 32m links. Source: /home/coolhand/html/datavis/data_trove/entertainment/movies/ml-32m/links.csv. Profiled with saturn-dissect v0.2.0 (saturn-insight-v2, anthropic:claude-opus-4-7). Retrieved from https://dr.eamer.dev/saturn/view/ml-32m-links