SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1901.02015 · 2019

Transit Least Squares: Optimized transit detection algorithm to search for periodic transits of small planets

arXiv · PDF · Open in the Atlas

Code that ran

8 of 12 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 0 failures of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1901.02015.json · badge: /api/ran/1901.02015.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
hippke/tls canonical (classifier label)
classifier basis: census
8 ran · 4 no recorded run
FunctionStatusWhere it lives
cleaned_array Ran
ran on a synthesized input
hippke/tls/transitleastsquares/helpers.py
code served (permissive licence) · get_code("5d99da9f9b486302")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy

# Create small numpy arrays with valid values
# The function expects arrays of measurements with time t, values y, and optionally errors dy

t = numpy.array([1.0, 2.0, 3.0, 4.0])
y = numpy.array([5.0, 10.0, 15.0, 20.0])
dy = numpy.array([0.5, 1.0, 1.5, 2.0])

ARGS = {
    "t": t,
    "y": y,
    "dy": dy
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
edge_effect_correction Ran
ran on a synthesized input
hippke/tls/transitleastsquares/core.py
code served (permissive licence) · get_code("bde3e43fa05e9011")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy
import numba

# Analyze the function to understand what's needed:
# - flux: used in (1 - flux) ** 2 * 1 / dy ** 2, so needs to be numeric array
# - patched_data: used in (1 - patched_data) ** 2 * inverse_squared_patched_dy
# - dy: used in 1 / dy ** 2, needs to be numeric array
# - inverse_squared_patched_dy: used directly in multiplication with (1 - patched_data) ** 2

# All parameters are used in numpy operations inside a numba.jit function,
# so they need to be numpy arrays.

# Create small numpy arrays with distinct dimensions for clarity
flux = numpy.array([0.1, 0.2, 0.3, 0.4], dtype=numpy.float64)
patched_data = numpy.array([0.15, 0.25, 0.35, 0.45], dtype=numpy.float64)
dy = numpy.array([1.0, 2.0, 3.0, 4.0], dtype=numpy.float64)
inverse_squared_patched_dy = numpy.array([1.0, 0.25, 0.111, 0.0625], dtype=numpy.float64)

ARGS = {
    "flux": flux,
    "patched_data": patched_data,
    "dy": dy,
    "inverse_squared_patched_dy": inverse_squared_patched_dy,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
fold Ran
ran on a synthesized input
hippke/tls/transitleastsquares/core.py
code served (permissive licence) · get_code("52c3d1f4ce1655b9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (5,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy
import numba

# The fold function performs phase folding with the following parameters:
# - time: array or scalar of time values
# - period: the period to fold by
# - T0: reference time (epoch)

# All parameters are numeric, so we'll create simple numpy arrays

time = numpy.array([0.0, 0.5, 1.0, 1.5, 2.0])
period = numpy.array(1.0)
T0 = numpy.array(0.0)

ARGS = {
    "time": time,
    "period": period,
    "T0": T0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (5,) — shape (5,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
foldfast Ran
ran on a synthesized input
hippke/tls/transitleastsquares/core.py
code served (permissive licence) · get_code("926863e948259e7d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (5,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy
import numba

# Create small arrays for the time and period parameters
time_array = numpy.array([0.1, 0.5, 1.2, 2.3, 3.7], dtype=numpy.float64)
period_value = numpy.float64(2.0)

ARGS = {
    "time": time_array,
    "period": period_value,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (5,) — shape (5,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
interpolation_search Ran
ran on a synthesized input
hippke/tls/transitleastsquares/interpolation.py
code served (permissive licence) · get_code("2cc07f5a00ed1299")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int, ndim 0, dtype family int, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numba
from math import floor
import numpy as np

# The function takes two parameters:
# - x: a sequence that will be indexed (needs len() and indexing with [])
# - z: a scalar value to search for

# Looking at the function body:
# - x needs to support len(), indexing with [], and comparison operations
# - z needs to support comparison operations with elements of x
# - x must be sorted in ascending order for the algorithm to work correctly
# - x should have at least 2 elements (assert n > 1)

# Create a sorted numpy array for x
x = np.array([1.0, 2.5, 4.0, 5.5, 7.0, 8.5, 10.0], dtype=np.float64)

# Create a scalar value z that falls within the range
z = 5.0

ARGS = {
    "x": x,
    "z": z
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
lerp Ran
ran on a synthesized input
hippke/tls/transitleastsquares/interpolation.py
code served (permissive licence) · get_code("1fa15722e6c09927")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numba
import numpy as np

# The function lerp takes two parameters:
# - y: an array-like object that supports indexing with [..., 0] and [..., 1]
# - theta: a scalar or array that can be broadcast with y[..., 0] and y[..., 1]

# Create a small numpy array with shape (2, 4, 2) to support the [..., 0] and [..., 1] indexing
# The last dimension has size 2, representing the two values to interpolate between
y = np.array([
    [[1.0, 2.0], [3.0, 4.0], [5.0, 6.0], [7.0, 8.0]],
    [[9.0, 10.0], [11.0, 12.0], [13.0, 14.0], [15.0, 16.0]]
], dtype=np.float64)

# theta should be a scalar or broadcastable value for interpolation parameter
theta = 0.5

ARGS = {
    "y": y,
    "theta": theta
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
pink_noise Ran
ran on a synthesized input
hippke/tls/transitleastsquares/stats.py
code served (permissive licence) · get_code("fa0f5b9a94285047")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float64, ndim 0, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy

# Construct valid arguments for pink_noise
# The function expects:
# - data: a sequence (list or array-like) that supports len() and slicing
# - width: an integer representing the window width

# Create a simple numpy array as data
data = numpy.array([1.0, 2.0, 3.0, 5.0, 4.0, 6.0, 7.0, 8.0, 9.0, 10.0])

# Create a reasonable width value (smaller than data length)
width = 3

ARGS = {
    "data": data,
    "width": width
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
transit_mask Ran
ran on a synthesized input
hippke/tls/transitleastsquares/helpers.py
code served (permissive licence) · get_code("60d51828dfc721d9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (6,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy

# The function performs modular arithmetic operations on numeric values
# t: time values (can be array-like)
# period: orbital period (scalar)
# duration: transit duration (scalar)
# T0: transit center time (scalar)

# Create small test arrays for the parameters
t = numpy.array([0.0, 0.5, 1.0, 1.5, 2.0, 2.5])
period = 2.0
duration = 0.4
T0 = 1.0

ARGS = {
    "t": t,
    "period": period,
    "duration": duration,
    "T0": T0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (6,) — shape (6,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
catalog_info_EPIC Not yet run
unverified (harvested, no recorded run)
hippke/tls/transitleastsquares/catalog.py
code served (permissive licence) · get_code("d203948861823901")
How it ran

no invocation recorded.

catalog_info_KIC Not yet run
unverified (harvested, no recorded run)
hippke/tls/transitleastsquares/catalog.py
code served (permissive licence) · get_code("56b9758a2f6351e0")
How it ran

no invocation recorded.

catalog_info_TIC Not yet run
unverified (harvested, no recorded run)
hippke/tls/transitleastsquares/catalog.py
code served (permissive licence) · get_code("1de64efa31f11620")
How it ran

no invocation recorded.

rp_rs_from_depth Not yet run
unverified (harvested, no recorded run)
hippke/tls/transitleastsquares/stats.py
code served (permissive licence) · get_code("1db54012152e965a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 12 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We present a new method to detect planetary transits from time-series photometry, the Transit Least Squares (TLS) algorithm. TLS searches for transit-like features while taking the stellar limb darkening and planetary ingress and egress into account. We have optimized TLS for both signal detection efficiency (SDE) of small planets and computational speed. TLS analyses the entire, unbinned phase-folded light curve. We compensate for the higher computational load by (i.) using algorithms like "Mergesort" (for the trial orbital phases) and by (ii.) restricting the trial transit durations to a smaller range that encompasses all known planets, and using stellar density priors where available. A typical K2 light curve, including 80d of observations at a cadence of 30min, can be searched with TLS in ~10s real time on a standard laptop computer, as fast as the widely used Box Least Squares (BLS) algorithm. We perform a transit injection-retrieval experiment of Earth-sized planets around sun-like stars using synthetic light curves with 110ppm white noise per 30min cadence, corresponding to a photometrically quiet KP=12 star observed with Kepler. We determine the SDE thresholds for both BLS and TLS to reach a false positive rate of 1% to be SDE~7 in both cases. The resulting true positive (or recovery) rates are ~93% for TLS and ~76% for BLS, implying more reliable detections with TLS. We also test TLS with the K2 light curve of the TRAPPIST-1 system and find six of seven Earth-sized planets using an iterative search for increasingly lower signal detection efficiency, the phase-folded transit of the seventh planet being affected by a stellar flare. TLS is more reliable than BLS in finding any kind of transiting planet but it is particularly suited for the detection of small planets in long time series from Kepler, TESS, and PLATO. We make our Python implementation of TLS publicly available.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("1901.02015")
get_code_for_paper("1901.02015")
have("1901.02015")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/1901.02015.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/1901.02015.svg)](https://syntology.ai/paper/1901.02015)

Connect an agent — have() is free.