mapFolding 0.15.4__tar.gz → 0.16.0__tar.gz

{mapfolding-0.15.4 → mapfolding-0.16.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mapFolding
-Version: 0.15.4
+Version: 0.16.0
 Summary: Map folding, meanders, stamp folding, semi-meanders. Experiment with algorithm transformations and code optimization.
 Author-email: Hunter Hogan <HunterHogan@pm.me>
 License: CC-BY-NC-4.0
@@ -8,7 +8,7 @@ Project-URL: Donate, https://www.patreon.com/integrated
 Project-URL: Homepage, https://github.com/hunterhogan/mapFolding
 Project-URL: Issues, https://github.com/hunterhogan/mapFolding/issues
 Project-URL: Repository, https://github.com/hunterhogan/mapFolding.git
-Keywords: A000136,A000560,A000682,A001010,A001011,A001415,A001416,A001417,A001418,A005315,A005316,A007822,A195646,A223094,A259702,A301620,AST manipulation,GPU acceleration,JIT compilation,Numba optimization,OEIS,Python optimization,abstract syntax tree,algorithmic combinatorics,algorithmic optimization,arch configurations,automated code generation,bit-packed arrays,bitwise state machines,cache-efficient algorithms,closed meandric numbers,code generation,code optimization,code synthesis,code transformation,codon optimization,combinatorial computing,combinatorial enumeration,combinatorial geometry,combinatorial mathematics,combinatorial problem solver,combinatorics,computational combinatorics,computational geometry,crossing patterns,curve crossings,dataclass transformation,discrete mathematics,dynamic compilation,enumerative combinatorics,folding pattern enumeration,folding problems,high-performance computing,integer sequences,just-in-time compilation,kernel optimization,labeled stamp folding,low-level computation,map folding,mapFolding,mathematical algorithms,mathematical modeling,mathematical optimization,mathematical patterns,mathematical software,mathematical tool,mathematical visualization,meander enumeration,meanders,meandric systems,memory-efficient enumeration,metaprogramming,numerical algorithms,numerical computation,open meandric systems,paper folding mathematics,parallel computing,pattern recognition,performance optimization,permutation patterns,permutations,post-setup optimization,pyproject,scientific computing,semi-meanders,sequence analysis,sequence calculator,sequence enumeration,sequence explorer,sequence generation,source code analysis,stamp folding,symbolic computation,symmetric foldings,topological combinatorics,topological patterns,typed Python
+Keywords: A000136,A000560,A000682,A001010,A001011,A001415,A001416,A001417,A001418,A005315,A005316,A007822,A178961,A195646,A223094,A259702,A301620,AST manipulation,GPU acceleration,JIT compilation,Numba optimization,OEIS,Python optimization,abstract syntax tree,algorithmic combinatorics,algorithmic optimization,arch configurations,automated code generation,bit-packed arrays,bitwise state machines,cache-efficient algorithms,closed meandric numbers,code generation,code optimization,code synthesis,code transformation,codon optimization,combinatorial computing,combinatorial enumeration,combinatorial geometry,combinatorial mathematics,combinatorial problem solver,combinatorics,computational combinatorics,computational geometry,crossing patterns,curve crossings,dataclass transformation,discrete mathematics,dynamic compilation,enumerative combinatorics,folding pattern enumeration,folding problems,high-performance computing,integer sequences,just-in-time compilation,kernel optimization,labeled stamp folding,low-level computation,map folding,mapFolding,mathematical algorithms,mathematical modeling,mathematical optimization,mathematical patterns,mathematical software,mathematical tool,mathematical visualization,meander enumeration,meanders,meandric systems,memory-efficient enumeration,metaprogramming,numerical algorithms,numerical computation,open meandric systems,paper folding mathematics,parallel computing,pattern recognition,performance optimization,permutation patterns,permutations,post-setup optimization,pyproject,scientific computing,semi-meanders,sequence analysis,sequence calculator,sequence enumeration,sequence explorer,sequence generation,source code analysis,stamp folding,symbolic computation,symmetric foldings,topological combinatorics,topological patterns,typed Python
 Classifier: Development Status :: 4 - Beta
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
@@ -35,6 +35,7 @@ Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: astToolkit
+Requires-Dist: autoflake
 Requires-Dist: hunterMakesPy
 Requires-Dist: numpy
 Requires-Dist: platformdirs
@@ -47,8 +48,14 @@ Requires-Dist: setuptools-scm; extra == "development"
 Provides-Extra: numba
 Requires-Dist: numba; extra == "numba"
 Requires-Dist: numba_progress; extra == "numba"
+Provides-Extra: pandas
+Requires-Dist: pandas; extra == "pandas"
+Requires-Dist: pyarrow; extra == "pandas"
+Requires-Dist: pyarrow-stubs; extra == "pandas"
 Provides-Extra: testing
 Requires-Dist: numba; extra == "testing"
+Requires-Dist: pandas; extra == "testing"
+Requires-Dist: pyarrow; extra == "testing"
 Requires-Dist: pytest-cov; extra == "testing"
 Requires-Dist: pytest-env; extra == "testing"
 Requires-Dist: pytest-xdist; extra == "testing"

{mapfolding-0.15.4 → mapfolding-0.16.0}/mapFolding/__init__.py

@@ -57,8 +57,7 @@ from mapFolding._theTypes import (
 	NumPyElephino as NumPyElephino,
 	NumPyFoldsTotal as NumPyFoldsTotal,
 	NumPyIntegerType as NumPyIntegerType,
-	NumPyLeavesTotal as NumPyLeavesTotal,
-)
+	NumPyLeavesTotal as NumPyLeavesTotal)
 
 from mapFolding._theSSOT import packageSettings as packageSettings
 
@@ -68,18 +67,18 @@ from mapFolding.beDRY import (
 	getTaskDivisions as getTaskDivisions,
 	makeDataContainer as makeDataContainer,
 	setProcessorLimit as setProcessorLimit,
-	validateListDimensions as validateListDimensions,
-)
+	validateListDimensions as validateListDimensions)
 
-from mapFolding.dataBaskets import MapFoldingState as MapFoldingState
+from mapFolding.dataBaskets import (
+    MapFoldingState as MapFoldingState,
+    MatrixMeandersState as MatrixMeandersState)
 
 from mapFolding.filesystemToolkit import (
 	getFilenameFoldsTotal as getFilenameFoldsTotal,
 	getPathFilenameFoldsTotal as getPathFilenameFoldsTotal,
 	getPathRootJobDEFAULT as getPathRootJobDEFAULT,
 	saveFoldsTotal as saveFoldsTotal,
-	saveFoldsTotalFAILearly as saveFoldsTotalFAILearly,
-)
+	saveFoldsTotalFAILearly as saveFoldsTotalFAILearly)
 
 from mapFolding.basecamp import countFolds as countFolds
 
@@ -89,5 +88,4 @@ from mapFolding.oeis import (
 	getFoldsTotalKnown as getFoldsTotalKnown,
 	getOEISids as getOEISids,
 	OEIS_for_n as OEIS_for_n,
-	oeisIDfor_n as oeisIDfor_n,
-)
+	oeisIDfor_n as oeisIDfor_n)

{mapfolding-0.15.4 → mapfolding-0.16.0}/mapFolding/_theSSOT.py

@@ -108,6 +108,7 @@ OEISidMeandersManuallySet: dict[str, MetadataOEISidMeandersManuallySet] = {
 	'A060206': {'valuesTestValidation': [*range(3, 9)]},
 	'A077460': {'valuesTestValidation': [*range(3, 8)]},
 	'A078591': {'valuesTestValidation': [*range(3, 10)]},
+	'A178961': {'valuesTestValidation': [*range(3, 11)]},
 	'A223094': {'valuesTestValidation': [*range(3, 11)]},
 	'A259702': {'valuesTestValidation': [*range(3, 13)]},
 	'A301620': {'valuesTestValidation': [*range(3, 11)]},

{mapfolding-0.15.4 → mapfolding-0.16.0}/mapFolding/algorithms/daoOfMapFolding.py

@@ -1,4 +1,3 @@
-# ruff: noqa: D100 D103
 from mapFolding.dataBaskets import MapFoldingState
 
 def activeLeafGreaterThan0(state: MapFoldingState) -> bool:
@@ -140,4 +139,4 @@ def count(state: MapFoldingState) -> MapFoldingState:
 
 def doTheNeedful(state: MapFoldingState) -> MapFoldingState:
 	state = count(state)
-	return state  # noqa: RET504
+	return state

mapfolding-0.16.0/mapFolding/algorithms/getBucketsTotal.py

@@ -0,0 +1,137 @@
+"""Buckets."""
+from mapFolding import MatrixMeandersState
+from mapFolding.reference.A005316facts import bucketsIf_k_EVEN_by_nLess_k, bucketsIf_k_ODD_by_nLess_k
+from math import exp, log
+from typing import NamedTuple
+import math
+
+class ImaKey(NamedTuple):
+	"""keys for dictionaries."""
+
+	oeisID: str
+	kIsOdd: bool
+	nLess_kIsOdd: bool
+
+def getBucketsTotal(state: MatrixMeandersState, safetyMultiplicand: float = 1.2) -> int:
+	"""Estimate the total number of non-unique curveLocations that will be computed from the existing curveLocations.
+
+	Notes
+	-----
+	Subexponential bucketsTotal unified estimator parameters (derived in reference notebook).
+
+	The model is: log(buckets) = intercept + bN*log(n) + bK*log(k) + bD*log(n-k) + g_r*(k/n) + g_r2*(k/n)^2 + g_s*((n-k)/n) + offset(subseries)
+	Subseries key: f"{oeisID}_kOdd={int(kIsOdd)}_dOdd={int(nLess_kIsOdd)}" with a reference subseries offset of zero.
+	These coefficients intentionally remain in-source (SSOT) to avoid runtime JSON parsing overhead and to support reproducibility.
+	"""
+	dictionaryExponentialCoefficients: dict[ImaKey, float] = {
+		(ImaKey(oeisID='', kIsOdd=False, nLess_kIsOdd=True)): 0.834,
+		(ImaKey(oeisID='', kIsOdd=False, nLess_kIsOdd=False)): 1.5803,
+		(ImaKey(oeisID='', kIsOdd=True, nLess_kIsOdd=True)): 1.556,
+		(ImaKey(oeisID='', kIsOdd=True, nLess_kIsOdd=False)): 1.8047,
+	}
+
+	logarithmicOffsets: dict[ImaKey, float] ={
+		(ImaKey('A000682', kIsOdd=False, nLess_kIsOdd=False)): 0.0,
+		(ImaKey('A000682', kIsOdd=False, nLess_kIsOdd=True)): -0.07302547148212568,
+		(ImaKey('A000682', kIsOdd=True, nLess_kIsOdd=False)): -0.00595307513938792,
+		(ImaKey('A000682', kIsOdd=True, nLess_kIsOdd=True)): -0.012201222865243722,
+		(ImaKey('A005316', kIsOdd=False, nLess_kIsOdd=False)): -0.6392728422078733,
+		(ImaKey('A005316', kIsOdd=False, nLess_kIsOdd=True)): -0.6904925299923548,
+		(ImaKey('A005316', kIsOdd=True, nLess_kIsOdd=False)): 0.0,
+		(ImaKey('A005316', kIsOdd=True, nLess_kIsOdd=True)): 0.0,
+	}
+
+	logarithmicParameters: dict[str, float] = {
+		'intercept': -166.1750299793178,
+		'log(n)': 1259.0051001675547,
+		'log(k)': -396.4306071056408,
+		'log(nLess_k)': -854.3309503739766,
+		'k/n': 716.530410654819,
+		'(k/n)^2': -2527.035113444166,
+		'normalized k': -882.7054406339189,
+	}
+
+	bucketsTotalMaximumBy_kOfMatrix: dict[int, int] = {1:3, 2:12, 3:40, 4:125, 5:392, 6:1254, 7:4087, 8:13623, 9:46181, 10:159137, 11:555469, 12:1961369, 13:6991893, 14:25134208}
+
+	xCommon = 1.57
+
+	nLess_k: int = state.n - state.kOfMatrix
+	kIsOdd: bool = bool(state.kOfMatrix & 1)
+	nLess_kIsOdd: bool = bool(nLess_k & 1)
+	kIsEven: bool = not kIsOdd
+	bucketsTotal: int = -8
+
+	"""NOTE temporary notes
+	I have a fault in my thinking. bucketsTotal increases as k decreases until ~0.4k, then bucketsTotal decreases rapidly to 1. I
+	have ignored the decreasing side. In the formulas for estimation, I didn't differentiate between increasing and decreasing.
+	So, I probably need to refine the formulas. I guess I need to add checks to the if/else monster, too.
+
+	While buckets is increasing:
+		3 types of estimates:
+			1. Exponential growth.
+			2. Logarithmic growth.
+			3. Hard ceiling.
+	While buckets is decreasing:
+		1. Hard ceiling, same as increasing side.
+		2. ???
+		3. buckets = 1.
+
+	The formula for exponential growth _never_ underestimates. I haven't measured by how much it overestimates.
+
+	"""
+
+	bucketsTotalAtMaximum: bool = state.kOfMatrix <= ((state.n - 1 - (state.kOfMatrix % 2)) // 3)
+	bucketsTotalGrowsExponentially: bool = state.kOfMatrix > nLess_k
+	bucketsTotalGrowsLogarithmically: bool = state.kOfMatrix > ((state.n - (state.n % 3)) // 3)
+
+	if bucketsTotalAtMaximum:
+		if state.kOfMatrix in bucketsTotalMaximumBy_kOfMatrix:
+			bucketsTotal = bucketsTotalMaximumBy_kOfMatrix[state.kOfMatrix]
+		else:
+			c = 0.95037
+			r = 3.3591258254
+			if kIsOdd:
+				c = 0.92444
+				r = 3.35776
+			bucketsTotal = int(c * r**state.kOfMatrix * safetyMultiplicand)
+
+	elif bucketsTotalGrowsExponentially:
+		if (state.oeisID == 'A005316') and kIsOdd and (nLess_k in bucketsIf_k_ODD_by_nLess_k):
+			# If I already know bucketsTotal.
+			bucketsTotal = bucketsIf_k_ODD_by_nLess_k[nLess_k]
+		elif (state.oeisID == 'A005316') and kIsEven and (nLess_k in bucketsIf_k_EVEN_by_nLess_k):
+			# If I already know bucketsTotal.
+			bucketsTotal = bucketsIf_k_EVEN_by_nLess_k[nLess_k]
+		else: # I estimate bucketsTotal during exponential growth.
+			xInstant: int = math.ceil(nLess_k / 2)
+			A000682adjustStartingCurveLocations: float = 0.25
+			startingConditionsCoefficient: float = dictionaryExponentialCoefficients[ImaKey('', kIsOdd, nLess_kIsOdd)]
+			if kIsOdd and nLess_kIsOdd:
+				A000682adjustStartingCurveLocations = 0.0
+			if state.oeisID == 'A000682': # NOTE Net effect is between `*= n` and `*= n * 2.2` if n=46.
+				startingConditionsCoefficient *= state.n * (((state.n // 2) + 2) ** A000682adjustStartingCurveLocations)
+			bucketsTotal = int(startingConditionsCoefficient * math.exp(xCommon * xInstant))
+
+	elif state.kOfMatrix <= max(bucketsTotalMaximumBy_kOfMatrix.keys()):
+		# If `kOfMatrix` is low, use maximum bucketsTotal. 1. Can't underestimate. 2. Skip computation that can underestimate.
+		# 3. The potential difference in memory use is relatively small.
+		bucketsTotal = bucketsTotalMaximumBy_kOfMatrix[state.kOfMatrix]
+
+	elif bucketsTotalGrowsLogarithmically:
+		xPower: float = (0
+			+ logarithmicParameters['intercept']
+			+ logarithmicParameters['log(n)'] * log(state.n)
+			+ logarithmicParameters['log(k)'] * log(state.kOfMatrix)
+			+ logarithmicParameters['log(nLess_k)'] * log(nLess_k)
+			+ logarithmicParameters['k/n'] * (state.kOfMatrix / state.n)
+			+ logarithmicParameters['(k/n)^2'] * (state.kOfMatrix / state.n)**2
+			+ logarithmicParameters['normalized k'] * nLess_k / state.n
+			+ logarithmicOffsets[ImaKey(state.oeisID, kIsOdd, nLess_kIsOdd)]
+		)
+
+		bucketsTotal = int(exp(xPower * safetyMultiplicand))
+
+	else:
+		message = "I shouldn't be here."
+		raise SystemError(message)
+	return bucketsTotal