juliensimon commited on
Commit
39941f7
·
verified ·
1 Parent(s): 2a4b187

Update meteorite database: 1,180 meteorites

Browse files
Files changed (3) hide show
  1. README.md +62 -57
  2. banner.jpg +3 -0
  3. data/meteorites.parquet +2 -2
README.md CHANGED
@@ -3,13 +3,7 @@ license: cc0-1.0
3
  pretty_name: "Meteorite Database"
4
  language:
5
  - en
6
- description: >-
7
- Known meteorites catalogued in Wikidata, including mass, classification,
8
- fall date, country of recovery, and geographic coordinates.
9
- 1,180 meteorites with metadata sourced from the community-curated
10
- Wikidata knowledge base.
11
- size_categories:
12
- - 1K<n<10K
13
  task_categories:
14
  - tabular-classification
15
  tags:
@@ -20,62 +14,74 @@ tags:
20
  - open-data
21
  - tabular-data
22
  - parquet
 
 
23
  configs:
24
  - config_name: default
25
- default: true
26
  data_files:
27
  - split: train
28
  path: data/meteorites.parquet
 
29
  ---
30
 
31
  # Meteorite Database
32
 
33
- *Part of the [Planetary Science Datasets](https://huggingface.co/collections/juliensimon/planetary-science-datasets-69c2d4683bd6a66c34fb4af2) collection on Hugging Face.*
34
 
35
- Catalogue of **1,180** known meteorites sourced from [Wikidata](https://www.wikidata.org/),
36
- covering mass, classification, fall date, country of recovery, and geographic coordinates.
 
 
 
 
 
37
 
38
  ## Dataset description
39
 
40
- Meteorites are extraterrestrial rocks that survive passage through Earth's atmosphere and
41
- reach the surface. They are classified by mineralogy and petrology (e.g., chondrites,
42
- achondrites, iron meteorites) and recorded either as *falls* (witnessed descent) or
43
- *finds* (recovered without observation).
 
 
44
 
45
- This dataset aggregates Wikidata entries for all entities of type Q60186 (meteorite), pulling
46
- structured properties including mass (P2067), fall/discovery date (P585/P575), country (P17),
47
- coordinates (P625), and mineralogical class (via P31 subclass hierarchy). It complements NASA and Meteoritical Society
48
- databases with Wikidata's multilingual, cross-linked knowledge graph.
49
 
50
  ## Schema
51
 
52
- | Column | Type | Description |
53
- |--------|------|-------------|
54
- | `wikidata_id` | string | Wikidata entity ID (e.g. Q1029) |
55
- | `name` | string | Meteorite name |
56
- | `fall_date` | string | Date of fall or recovery (YYYY-MM-DD) |
57
- | `mass_g` | float | Mass in grams |
58
- | `classification` | string | Mineralogical/petrological classification |
59
- | `country` | string | Country of recovery |
60
- | `latitude` | float | Recovery latitude (decimal degrees) |
61
- | `longitude` | float | Recovery longitude (decimal degrees) |
62
 
63
  ## Quick stats
64
 
65
  - **1,180** meteorites total
66
  - **368** with recorded mass
67
- - **415** with geographic coordinates
68
  - **125** with classification
69
  - **55** countries of recovery
70
  - Heaviest: Allende meteorite (2,000,000 g)
71
  - Top countries: United States (39), Australia (13), Argentina (10), Germany (10), France (9)
72
- - Top classifications: iron meteorite (21), chondrite (17), H chondrite (17), L chondrite (11), ordinary chondrite (11)
73
 
74
  ## Usage
75
 
76
  ```python
77
  from datasets import load_dataset
78
 
 
 
 
 
 
 
 
79
  ds = load_dataset("juliensimon/meteorite-database", split="train")
80
  df = ds.to_pandas()
81
 
@@ -83,53 +89,52 @@ df = ds.to_pandas()
83
  print(df.nlargest(10, "mass_g")[["name", "mass_g", "country", "classification"]])
84
 
85
  # Meteorites by country
86
- print(df["country"].value_counts().head(10))
87
-
88
- # Meteorites with coordinates (mappable)
89
- mappable = df.dropna(subset=["latitude", "longitude"])
90
- print(f"{len(mappable):,} meteorites with coordinates")
91
-
92
- # Filter by classification
93
- chondrites = df[df["classification"].str.contains("chondrite", case=False, na=False)]
94
- print(f"{len(chondrites):,} chondrites")
 
 
 
 
 
 
 
95
  ```
96
 
97
  ## Data source
98
 
99
- [Wikidata](https://www.wikidata.org/) SPARQL endpoint. Meteorites identified via
100
- property P31 (instance of) = Q60186 (meteorite). Data is community-curated and
101
- cross-referenced with the [Meteoritical Bulletin Database](https://www.lpi.usra.edu/meteor/).
102
-
103
- ## Update schedule
104
-
105
- Quarterly (January, April, July, October). Run manually to capture interim additions.
106
 
107
  ## Related datasets
108
 
109
- - [impact-craters](https://huggingface.co/datasets/juliensimon/impact-craters) -- Earth impact crater database
110
- - [fireballs](https://huggingface.co/datasets/juliensimon/fireball-bolide-events) -- NASA fireball and bolide events
111
 
112
- ## Pipeline
113
 
114
- Source code: [juliensimon/space-datasets](https://github.com/juliensimon/space-datasets)
115
 
116
- ## Support
117
 
118
- If you find this dataset useful, please give it a ❤️ on the [dataset page](https://huggingface.co/datasets/juliensimon/meteorite-database) and share feedback in the Community tab! Also consider giving a ⭐️ to the [space-datasets](https://github.com/juliensimon/space-datasets) repo.
119
 
120
  ## Citation
121
 
122
  ```bibtex
123
  @dataset{meteorite_database,
124
- author = {Simon, Julien},
125
  title = {Meteorite Database},
 
126
  year = {2026},
127
- publisher = {Hugging Face},
128
  url = {https://huggingface.co/datasets/juliensimon/meteorite-database},
129
- note = {Sourced from Wikidata (CC0)}
130
  }
131
  ```
132
 
133
  ## License
134
 
135
- [CC0-1.0](https://creativecommons.org/publicdomain/zero/1.0/) (Wikidata content is public domain)
 
3
  pretty_name: "Meteorite Database"
4
  language:
5
  - en
6
+ description: "Catalogue of known meteorites sourced from Wikidata, covering mass, classification, fall date, country of recovery, and geographic coordinates. Meteorites are extraterrestrial rocks that survive pass"
 
 
 
 
 
 
7
  task_categories:
8
  - tabular-classification
9
  tags:
 
14
  - open-data
15
  - tabular-data
16
  - parquet
17
+ size_categories:
18
+ - 1K<n<10K
19
  configs:
20
  - config_name: default
 
21
  data_files:
22
  - split: train
23
  path: data/meteorites.parquet
24
+ default: true
25
  ---
26
 
27
  # Meteorite Database
28
 
 
29
 
30
+ <div align="center">
31
+ <img src="banner.jpg" alt="Rosetta spacecraft approaching Comet 67P/Churyumov-Gerasimenko" width="400">
32
+ <p><em>Credit: NASA/ESA</em></p>
33
+ </div>
34
+
35
+
36
+ *Part of a [dataset collection](https://huggingface.co/collections/juliensimon/orbital-mechanics-datasets-69c24caca4ab3934c9856994) on Hugging Face.*
37
 
38
  ## Dataset description
39
 
40
+ Catalogue of known meteorites sourced from Wikidata, covering mass, classification, fall date, country of recovery, and geographic coordinates.
41
+
42
+ Meteorites are extraterrestrial rocks that survive passage through Earth's atmosphere and reach the surface. They are classified by mineralogy and petrology (e.g., chondrites, achondrites, iron meteorites) and recorded either as falls (witnessed descent) or finds (recovered without observation).
43
+
44
+ This dataset aggregates Wikidata entries for all entities of type Q60186 (meteorite), pulling structured properties including mass (P2067), fall/discovery date (P585/P575), country (P17), coordinates (P625), and mineralogical class (via P31 subclass hierarchy). It complements NASA and Meteoritical Society databases with Wikidata's multilingual, cross-linked knowledge graph.
45
+
46
 
47
+ This dataset is suitable for **tabular classification** tasks.
 
 
 
48
 
49
  ## Schema
50
 
51
+ | Column | Type | Description | Sample | Null % |
52
+ |--------|------|-------------|--------|--------|
53
+ | `wikidata_id` | str | Wikidata entity ID (e.g. Q1029); stable cross-reference key for linking to other Wikidata properties | Q138638076 | 0.0% |
54
+ | `name` | string | Official meteorite name assigned by the Meteoritical Society (e.g., 'Allende', 'NWA 7034', 'Chelyabinsk'); typically location of find plus sequence number | 2026 Koblenz meteor | 0.0% |
55
+ | `fall_date` | str | Date of observed fall or discovery/recovery in ISO format (YYYY-MM-DD); null for historical finds without a recorded date; precision often year-only (day defaults to 01) | 1880-01-01 | 53.8% |
56
+ | `mass_g` | float64 | Total known mass in grams; null if unknown; range from <1 g (tiny fragments) to ~60,000,000 g (Hoba, the largest known meteorite) | 21.0 | 68.8% |
57
+ | `classification` | string | Meteoritical Society mineralogical/petrological class (e.g., 'L5', 'CM2', 'Iron IIIAB'); letters = chemical group, numbers = petrologic grade; null if not recorded in Wikidata | ordinary chondrite | 89.4% |
58
+ | `country` | string | Country of recovery (English label from Wikidata); null for finds without a recorded country or in international territory (e.g., Antarctica) | Denmark | 83.0% |
59
+ | `latitude` | float64 | Recovery location latitude in decimal degrees (positive = N, negative = S); null for historical or poorly documented finds | 50.775 | 64.9% |
60
+ | `longitude` | float64 | Recovery location longitude in decimal degrees (positive = E, negative = W); null for historical or poorly documented finds | 6.08333 | 64.9% |
61
 
62
  ## Quick stats
63
 
64
  - **1,180** meteorites total
65
  - **368** with recorded mass
66
+ - **414** with geographic coordinates
67
  - **125** with classification
68
  - **55** countries of recovery
69
  - Heaviest: Allende meteorite (2,000,000 g)
70
  - Top countries: United States (39), Australia (13), Argentina (10), Germany (10), France (9)
71
+ - Top classifications: iron meteorite (20), H chondrite (17), chondrite (15), ordinary chondrite (14), L chondrite (10)
72
 
73
  ## Usage
74
 
75
  ```python
76
  from datasets import load_dataset
77
 
78
+ ds = load_dataset("juliensimon/meteorite-database", split="train")
79
+ df = ds.to_pandas()
80
+ ```
81
+
82
+ ```python
83
+ from datasets import load_dataset
84
+
85
  ds = load_dataset("juliensimon/meteorite-database", split="train")
86
  df = ds.to_pandas()
87
 
 
89
  print(df.nlargest(10, "mass_g")[["name", "mass_g", "country", "classification"]])
90
 
91
  # Meteorites by country
92
+ import matplotlib.pyplot as plt
93
+ df["country"].value_counts().head(15).plot.barh()
94
+ plt.xlabel("Count")
95
+ plt.ylabel("Country")
96
+ plt.title("Meteorites by Country of Recovery")
97
+ plt.tight_layout()
98
+ plt.show()
99
+
100
+ # Mass distribution (log scale)
101
+ import numpy as np
102
+ masses = df["mass_g"].dropna()
103
+ plt.hist(np.log10(masses[masses > 0]), bins=50)
104
+ plt.xlabel("log10(mass in grams)")
105
+ plt.ylabel("Count")
106
+ plt.title("Meteorite Mass Distribution")
107
+ plt.show()
108
  ```
109
 
110
  ## Data source
111
 
112
+ https://www.wikidata.org/
 
 
 
 
 
 
113
 
114
  ## Related datasets
115
 
116
+ - [juliensimon/impact-craters](https://huggingface.co/datasets/juliensimon/impact-craters)
 
117
 
118
+ - [juliensimon/fireball-bolide-events](https://huggingface.co/datasets/juliensimon/fireball-bolide-events)
119
 
120
+ > If you find this dataset useful, please consider [giving it a like](https://huggingface.co/datasets/juliensimon/meteorite-database) on Hugging Face. It helps others discover it.
121
 
122
+ ## About the author
123
 
124
+ Created by [Julien Simon](https://julien.org) AI Operating Partner at Fortino Capital. Part of the [Space Datasets](https://julien.org/datasets) collection.
125
 
126
  ## Citation
127
 
128
  ```bibtex
129
  @dataset{meteorite_database,
 
130
  title = {Meteorite Database},
131
+ author = {juliensimon},
132
  year = {2026},
 
133
  url = {https://huggingface.co/datasets/juliensimon/meteorite-database},
134
+ publisher = {Hugging Face}
135
  }
136
  ```
137
 
138
  ## License
139
 
140
+ [CC0-1.0](https://creativecommons.org/licenses/by/4.0/)
banner.jpg ADDED

Git LFS Details

  • SHA256: 9aaf8dfe42e887d82d0642bb2bfc68d3ea0e210eaa3f0f2cb276f35730922fb5
  • Pointer size: 130 Bytes
  • Size of remote file: 24.9 kB
data/meteorites.parquet CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b1e8fe5f819ab7bbdb7face41acefbfc1d649fa8f2c4b5d95b3ffd3e271764df
3
- size 32076
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1968a1cc33fb33fadbb39e6caca864f4ba6d5ebb46b8320999cbc4e9a5b0060e
3
+ size 32034