Dataset Viewer
Auto-converted to Parquet Duplicate
date
stringdate
1900-01-01 00:00:00
2099-12-31 00:00:00
day_reduced
int64
1
22
month_reduced
int64
1
11
year_reduced
int64
1
22
parts_sum
int64
3
55
life_path
int64
1
33
life_path_all_digits
int64
1
33
1900-01-01
1
1
1
3
3
3
1900-01-02
2
1
1
4
4
4
1900-01-03
3
1
1
5
5
5
1900-01-04
4
1
1
6
6
6
1900-01-05
5
1
1
7
7
7
1900-01-06
6
1
1
8
8
8
1900-01-07
7
1
1
9
9
9
1900-01-08
8
1
1
10
1
1
1900-01-09
9
1
1
11
11
2
1900-01-10
1
1
1
3
3
3
1900-01-11
11
1
1
13
4
4
1900-01-12
3
1
1
5
5
5
1900-01-13
4
1
1
6
6
6
1900-01-14
5
1
1
7
7
7
1900-01-15
6
1
1
8
8
8
1900-01-16
7
1
1
9
9
9
1900-01-17
8
1
1
10
1
1
1900-01-18
9
1
1
11
11
2
1900-01-19
1
1
1
3
3
3
1900-01-20
2
1
1
4
4
4
1900-01-21
3
1
1
5
5
5
1900-01-22
22
1
1
24
6
6
1900-01-23
5
1
1
7
7
7
1900-01-24
6
1
1
8
8
8
1900-01-25
7
1
1
9
9
9
1900-01-26
8
1
1
10
1
1
1900-01-27
9
1
1
11
11
2
1900-01-28
1
1
1
3
3
3
1900-01-29
11
1
1
13
4
22
1900-01-30
3
1
1
5
5
5
1900-01-31
4
1
1
6
6
6
1900-02-01
1
2
1
4
4
4
1900-02-02
2
2
1
5
5
5
1900-02-03
3
2
1
6
6
6
1900-02-04
4
2
1
7
7
7
1900-02-05
5
2
1
8
8
8
1900-02-06
6
2
1
9
9
9
1900-02-07
7
2
1
10
1
1
1900-02-08
8
2
1
11
11
2
1900-02-09
9
2
1
12
3
3
1900-02-10
1
2
1
4
4
4
1900-02-11
11
2
1
14
5
5
1900-02-12
3
2
1
6
6
6
1900-02-13
4
2
1
7
7
7
1900-02-14
5
2
1
8
8
8
1900-02-15
6
2
1
9
9
9
1900-02-16
7
2
1
10
1
1
1900-02-17
8
2
1
11
11
2
1900-02-18
9
2
1
12
3
3
1900-02-19
1
2
1
4
4
22
1900-02-20
2
2
1
5
5
5
1900-02-21
3
2
1
6
6
6
1900-02-22
22
2
1
25
7
7
1900-02-23
5
2
1
8
8
8
1900-02-24
6
2
1
9
9
9
1900-02-25
7
2
1
10
1
1
1900-02-26
8
2
1
11
11
2
1900-02-27
9
2
1
12
3
3
1900-02-28
1
2
1
4
4
22
1900-03-01
1
3
1
5
5
5
1900-03-02
2
3
1
6
6
6
1900-03-03
3
3
1
7
7
7
1900-03-04
4
3
1
8
8
8
1900-03-05
5
3
1
9
9
9
1900-03-06
6
3
1
10
1
1
1900-03-07
7
3
1
11
11
2
1900-03-08
8
3
1
12
3
3
1900-03-09
9
3
1
13
4
22
1900-03-10
1
3
1
5
5
5
1900-03-11
11
3
1
15
6
6
1900-03-12
3
3
1
7
7
7
1900-03-13
4
3
1
8
8
8
1900-03-14
5
3
1
9
9
9
1900-03-15
6
3
1
10
1
1
1900-03-16
7
3
1
11
11
2
1900-03-17
8
3
1
12
3
3
1900-03-18
9
3
1
13
4
22
1900-03-19
1
3
1
5
5
5
1900-03-20
2
3
1
6
6
6
1900-03-21
3
3
1
7
7
7
1900-03-22
22
3
1
26
8
8
1900-03-23
5
3
1
9
9
9
1900-03-24
6
3
1
10
1
1
1900-03-25
7
3
1
11
11
2
1900-03-26
8
3
1
12
3
3
1900-03-27
9
3
1
13
4
22
1900-03-28
1
3
1
5
5
5
1900-03-29
11
3
1
15
6
6
1900-03-30
3
3
1
7
7
7
1900-03-31
4
3
1
8
8
8
1900-04-01
1
4
1
6
6
6
1900-04-02
2
4
1
7
7
7
1900-04-03
3
4
1
8
8
8
1900-04-04
4
4
1
9
9
9
1900-04-05
5
4
1
10
1
1
1900-04-06
6
4
1
11
11
2
1900-04-07
7
4
1
12
3
3
1900-04-08
8
4
1
13
4
22
1900-04-09
9
4
1
14
5
5
1900-04-10
1
4
1
6
6
6
End of preview. Expand in Data Studio

Life Path number distribution over the 1900-2099 civil calendar

Every one of the 73,049 calendar dates from 1900-01-01 to 2099-12-31, each reduced to its Pythagorean Life Path number, plus the resulting frequency distribution.

License: Creative Commons Attribution 4.0 International (CC-BY-4.0) · Canonical source: https://numeroai.me/articles/life-path-number-distribution/ (RU)

What is in here

File Size What it is
life-path-dates.csv 1.8 MB Row-level table: one line per calendar date with every reduced component.
life-path-distribution.csv 0.2 KB Aggregate: dates per Life Path number and its share.
life-path-by-reduction-order.csv 0.4 KB The same census under both reduction orders, side by side.
life-path-distribution.json 3.7 KB Aggregate plus methodology, master-number siphon, rarest-number decomposition.
datapackage.json 4.4 KB Frictionless Data descriptor (field types and descriptions).
reproduce.py 7.4 KB Zero-dependency verifier: recomputes everything and asserts it matches.

Methodology

Life Path = reduce(day) + reduce(month) + reduce(digit-sum of year), then reduce the total. Master numbers 11, 22, 33 are preserved at every reduction step (Pythagorean school). No sampling, no rounding of counts: every date in the range is enumerated once and tallied once. The generator runs the same production code the calculator at numeroai.me uses (app/services/numerology/calculator.py :: life_path_number), so a regenerated dataset always matches what the product computes. The files here are a snapshot of that run, not a live query: the guarantee holds as of the last regeneration, and reproduce.py is what lets you confirm the snapshot you hold is internally consistent.

The generator is open and deterministic — rerun it and the counts match to the unit. The row-level table exists precisely so the aggregate does not have to be trusted: group life-path-dates.csv by life_path and you must reproduce life-path-distribution.csv exactly.

What the data shows

Digit-summing preserves a number modulo 9, so the Life Path is a mod-9 function of the date in a costume — and the master numbers are congruent to the digits they replace (11 to 2, 22 to 4, 33 to 6). Everything in the table follows from that. Six numbers (1, 3, 5, 7, 8, 9) sit within four dates of one ninth, 11.11% each (perfect ninths are impossible: 73,049 is not divisible by 9). The dips at 2 (5.00%), 4 (7.74%) and 6 (10.58%) are the masters taking part of their residue class, and each digit-plus-master pair adds back to one ninth. The rarest value is 33 — 393 dates, 0.54% — which also makes Life Path 2 (5.00%) rarer than the "rare" master 11 (6.11%).

Which reduction order? (this changes the answer)

Two conventions are in common use inside the same master-preserving school, and they disagree about how rare a number is:

33 22 11
reduce day, month, year separately, then add 393 (0.54%) 2,463 (3.37%) 4,464 (6.11%)
sum all eight digits of the date, then reduce 2,186 (2.99%) 3,782 (5.18%) 4,812 (6.59%)

That is a 5.6x swing on the headline number, and the two orders disagree on 10,374 of the 73,049 dates (14.2%) — the six numbers with no master in their residue class are unaffected. The main tables here use the first order (it is what the product computes); life-path-by-reduction-order.csv carries both, and life-path-dates.csv has a column for each. If you cite a rarity figure from this dataset, cite the reduction order with it — otherwise the figure means little.

Scope and limits (please read before citing)

  • This is a census of calendar dates, not of people. Real birthdays are not uniformly distributed across the calendar, so the share of living people holding each number differs from the share of dates.
  • Master numbers are preserved (the Pythagorean school). A school that reduces 11/22/33 to 2/4/6 would collapse this to nine rows — you can do that from the data above, since each master's dates simply return to 2, 4 or 6.
  • Shares are properties of the 1900-2099 window, not of "the calendar" in general. All seven year-surge years that feed 33 fall in the 20th century, so 33 covers 0.775% of 20th-century dates but only 0.301% of 21st-century ones.
  • The meanings attached to these numbers are a cultural tradition, not a scientific finding. The dataset describes the arithmetic of a reduction rule, nothing more.

Verify it yourself

reproduce.py has no dependencies. It recomputes all 73,049 dates from scratch and asserts every published figure against the shipped files — both reduction orders, every share, the siphon pairs, the master-number group, and the full decomposition of the rarest number. If any number here were wrong, it would exit non-zero.

python3 reproduce.py

How to cite

Znakai (2026). Life Path number distribution over the 1900-2099 civil calendar [Data set]. CC BY 4.0. https://numeroai.me/articles/life-path-number-distribution/

Related reading

Who publishes this

Znakai, the team behind the numerology apps at numeroai.me (EN) and ai-numerolog.ru (RU), which have a paid tier. We publish the arithmetic openly because the arithmetic is the only part of numerology that can be checked — and we would rather be the people who show the reduction steps than the people who do not. The data is yours under CC BY 4.0 regardless of what you think of the subject.

Downloads last month
27