1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
|
# Notes on MAG
Using: https://archive.org/details/mag-2021-06-07
* 260M entities
* 96.7M with DOI
In order to generate a doi-to-doi version, we need to:
* create a mapping from id-to-doi (db is slow, in-memory needs 16G+ RAM)
* apply the mapping to the PaperReferences file
```
# https://docs.microsoft.com/en-us/academic-services/graph/reference-data-schema
# MAG has a Papers.txt.gz file containing the ID and DOI
#
# (1) create a mapping from id to doi (e.g. in sqlite3)
# (2) turn id-to-id references into doi-to-doi with lookup table
#
# $ unpigz -c /magna/data/mag-2020-06-25/Papers.txt.gz | cut -f1,3 | pv -l
# 238M 0:11:16 [ 353k/s]
#
# $ unpigz -c /magna/data/mag-2021-06-07/Papers.txt.gz | cut -f1,3 | pv -l > /dev/null
# 260M 0:10:32 [ 412k/s]
#
# $ unpigz -c /magna/data/mag-2021-06-07/Papers.txt.gz | cut -f1,3 | awk '$2 != ""' | pv -l > /dev/null
# 96.7M 0:11:05 [ 145k/s]
#
# $ time unpigz -c /magna/data/mag-2021-06-07/Papers.txt.gz | cut -f1,3 | awk '$2 != ""' | mkocidb -o mag_id_doi.db
#
# 2021/09/27 17:08:45 [ok] initialized database -- /sandcrawler-db/tmp-refcat/mag_id_doi.db
# written 2.9G -- 3.4M/s
# 2021/09/27 17:23:11 import done
# 2021/09/27 17:23:11 creating index
# 2021/09/27 17:26:44 [ok] 1/2 created index -- /sandcrawler-db/tmp-refcat/mag_id_doi.db
# 2021/09/27 17:31:53 [ok] 2/2 created index -- /sandcrawler-db/tmp-refcat/mag_id_doi.db
#
# real 23m7.744s
# user 30m55.642s
# sys 4m17.959s
#
# Can use a in memory map, too - sqlite3 lookups for 2B+ items takes a while at
# 30 Kqps; takes 30% of RAM on 48GB, sigh; just map[int]string from paper id to doi.
#
# Prepare the 2-TSV list.
#
# $ time unpigz -c /magna/data/mag-2021-06-07/Papers.txt.gz | cut -f1,3 | awk '$2 != ""' | pv -l > /magna/data/mag-2021-06-07/mapping.tsv
#
# real 15m27.348s
# user 21m4.297s
# sys 3m34.441s
#
# $ time zstdcat -T0 PaperReferences.txt.zst |
# magrefs-mem -f /magna/data/mag-2021-06-07/mapping.tsv |
# pv -l | zstd -c -T0 > doi_refs.tsv.zst
#
# real 41m12.911s
# user 223m14.467s
# sys 30m43.260s
#
# $ zstdcat -T0 doi_refs.tsv.zst| pv -l | wc -l
# 1.32G 0:06:33 [3.34M/s] [ <=> ]1315040677
# 1315040677
```
Finding 1,315,040,677 DOI-to-DOI mappings.
Total edges: 1,832,226,781
```
$ zstdcat -T0 PaperReferences.txt.zst | pv -l | wc -l
1832226781
```
Non-DOI edges: 517,186,104
Creating lowercase, unique sorted version:
```
$ time zstdcat -T0 doi_refs.tsv.zst| tr '[[:upper:]]' '[[:lower:]]' | LC_ALL=C sort -u -T /sandcrawler-db/tmp-refcat/ -S50% > doi_refs_lower_sorted.tsv.zst
```
|