jkomoros/sudoku issues and pull requests

#315 - Use richer doc comments in go 1.19

Issue - State: open - Opened by jkomoros about 2 years ago

#314 - How can i run this Project

Issue - State: open - Opened by afkcodes over 2 years ago - 4 comments

#313 - Use context for deadlines in Techniques

Issue - State: open - Opened by jkomoros almost 7 years ago
Labels: sudoku

#312 - Use the Functional Options Pattern for HumanSolveOptions, GenerateOptions

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: cleanup

#311 - Investigate performance with Prism

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: performance

#310 - {Naked,Hidden}{Pair,Triple,Quad} Should be one technique with two variants

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: performance, humansolve

#309 - In analysis-pipeline, use os.Executable()

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: analysis, cleanup

#308 - ProbabilityDistribution.Invert should be weaker when all items are large

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: humansolve, calibration

#307 - Forcing Chains base difficulty is too high

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: calibration

#306 - Twiddlers have a very weak effect on very difficult techniques

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: humansolve

#305 - Case study: cull step not picked. Guesses sorted above others.

Issue - State: open - Opened by jkomoros over 7 years ago - 7 comments
Labels: bug, humansolve

#304 - Consider having more methods return CellRefs

Issue - State: open - Opened by jkomoros over 7 years ago

#303 - IntSlice.Intersection(other IntSlice) is a bottleneck

Issue - State: open - Opened by jkomoros over 7 years ago - 2 comments
Labels: performance, humansolve

#302 - CellSlice, MutableCellSlice, CellRefSlice have many similar methods

Issue - State: open - Opened by jkomoros over 7 years ago
Labels: cleanup

#301 - Get rid of _NUM_SOLVER_THREADS

Issue - State: open - Opened by jkomoros almost 8 years ago
Labels: cleanup

#300 - Load, LoadSDK, etc should be package level functions

Issue - State: closed - Opened by jkomoros almost 8 years ago

#299 - Add package level comments about Grid/MutableGrid, how to use the package, etc

Issue - State: open - Opened by jkomoros almost 8 years ago

#298 - Remove all private methods on Cell, MutableCell, Grid,MutableGrid interfaces

Issue - State: open - Opened by jkomoros almost 8 years ago - 1 comment

#297 - Consider having a package-public notion of CellRef and CellRefSlice

Issue - State: closed - Opened by jkomoros almost 8 years ago - 1 comment

#296 - Investigate interface fuzz testing for Grid/MutableGrid

Issue - State: open - Opened by jkomoros almost 8 years ago - 1 comment

#295 - Consider getting rid of HumanSolve, Solve and adding Grid.Replace

Issue - State: open - Opened by jkomoros almost 8 years ago
Labels: cleanup

#294 - Crazy idea: what if every CellSlice method on Grid returned a slice of cellrefs?

Issue - State: open - Opened by jkomoros almost 8 years ago
Labels: cleanup, performance

#293 - ChainedSteps should give SOME points if the column/row is off by one

Issue - State: closed - Opened by jkomoros almost 8 years ago - 1 comment
Labels: enhancement, analysis, humansolve

#292 - Forcing Chain is never picked against Guess

Issue - State: closed - Opened by jkomoros almost 8 years ago - 1 comment
Labels: bug, analysis, humansolve

#291 - Investigate a difficulty model using an alternate architecture

Issue - State: open - Opened by jkomoros almost 8 years ago
Labels: analysis, humansolve

#290 - Consider having humanLikelihood not be a twiddler

Issue - State: open - Opened by jkomoros almost 8 years ago
Labels: analysis, humansolve

#289 - i-sudoku should switch to default number of options to calculate

Issue - State: closed - Opened by jkomoros almost 8 years ago
Labels: cleanup, i-sudoku

#288 - The top choices in DebugHint are WAY too far in front of other techniques.

Issue - State: open - Opened by jkomoros about 8 years ago - 2 comments
Labels: analysis, humansolve, i-sudoku

#287 - Start tagging releases

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: cleanup

#286 - Create a ChainSteps boost for being the same number

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: humansolve

#285 - CommonNumbers twiddler should return DIM for numbers that aren't filled

Issue - State: closed - Opened by jkomoros about 8 years ago - 1 comment
Labels: bug, humansolve

#284 - The new CompoundSolveStep solver should reuse cull steps

Issue - State: open - Opened by jkomoros about 8 years ago - 1 comment
Labels: performance, humansolve

#283 - Chain Similarities Twiddler should have expontential backoff

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis, humansolve

#282 - Consider using more cores in dokugen-analysis

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: performance, humansolve

#281 - Look for performance improvements with profiling

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: performance, humansolve

#280 - Only Legal Number's humanLikelihood is too low

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: humansolve

#279 - If you do a step and undo it, debug hint seems to have forgotten the right thing for chain similarity

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: bug, i-sudoku

#278 - Pointing-Target Overlap way too strong

Issue - State: closed - Opened by jkomoros about 8 years ago - 2 comments
Labels: bug, humansolve, i-sudoku

#277 - i-sudoku hint doesn't appear to be selecting from the same distribution as ctrl-h

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: bug, i-sudoku

#276 - Add a way to attenuate the strength of each twiddler

Issue - State: closed - Opened by jkomoros about 8 years ago - 1 comment
Labels: enhancement, humansolve

#275 - ChainSimilarity Twiddler appears to be pushign numbers in the wrong direction

Issue - State: closed - Opened by jkomoros about 8 years ago - 1 comment
Labels: bug, analysis, humansolve

#274 - Remove HumanSolveOptions.effectiveTechniquesToUse

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: cleanup

#273 - Figure out a better way to get data to i-sudoku's debug-hint

Issue - State: open - Opened by jkomoros about 8 years ago - 1 comment
Labels: cleanup

#272 - Deeply investigate debug-hints in i-sudoku

Issue - State: open - Opened by jkomoros about 8 years ago - 2 comments
Labels: analysis, humansolve

#271 - In debug-hint view, allow picking which hint to apply

Issue - State: open - Opened by jkomoros about 8 years ago - 2 comments
Labels: enhancement, i-sudoku

#270 - Signal: Length of Longest non-fill run

Issue - State: closed - Opened by jkomoros about 8 years ago - 1 comment
Labels: enhancement, humansolve

#269 - Remove use of 'self' in method receivers

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: cleanup

#268 - Create a CellRefSlice?

Issue - State: closed - Opened by jkomoros about 8 years ago - 1 comment
Labels: sudoku, cleanup, performance

#267 - Cull steps tell users to cull possibilities from filled cells unnecessarily

Issue - State: open - Opened by jkomoros about 8 years ago - 3 comments
Labels: humansolve

#266 - Use structhash for the ids of the difficulty models

Issue - State: open - Opened by jkomoros about 8 years ago

#265 - Investigate a sentiment-analysis-like model

Issue - State: open - Opened by jkomoros about 8 years ago - 1 comment
Labels: analysis, humansolve

#264 - Get r2 back up to pre-auto-skew levels

Issue - State: open - Opened by jkomoros about 8 years ago - 2 comments
Labels: analysis, humansolve

#263 - Analysis-pipeline histogram should also spit out a picture

Issue - State: open - Opened by jkomoros about 8 years ago - 1 comment
Labels: enhancement, analysis

#262 - Allow ability to print histogram of relative_difficulties as well

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#261 - Histogram phase should be in the inner branch loop in analysis-pipeline

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#260 - Generating a medium or tough sudoku takes forever

Issue - State: closed - Opened by samifruit514 about 8 years ago - 3 comments
Labels: bug, analysis

#259 - an -overwrite arg to analysis-pipeline

Issue - State: open - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#258 - Various usability improvements to analysis-pipeline

Issue - State: open - Opened by jkomoros about 8 years ago - 1 comment
Labels: enhancement, analysis

#257 - analysis-pipeline should be configured into phases

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#256 - Build a tool to parse Weka output into a weights map and R2

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#255 - Allow loading a difficulty model at run time

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, humansolve

#254 - analysis-pipeline should be able to show a histogram of generated difficulties with that model

Issue - State: closed - Opened by jkomoros about 8 years ago
Labels: enhancement, analysis

#253 - Investigate why generated puzzles have a lower difficulty

Issue - State: closed - Opened by jkomoros over 8 years ago - 20 comments
Labels: bug, analysis, humansolve

#252 - komo format appears to be broken, exporting 3:3 instead of 3!

Issue - State: closed - Opened by jkomoros over 8 years ago - 2 comments
Labels: bug, sudoku

#251 - Merge weka-trainer into analysis-pipeline

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: analysis, cleanup

#250 - a-b-tester should acquire a wake lock when running

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#249 - a-b-tester should cleanup files even if closed early

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#248 - Rename a-b-tester

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: analysis, cleanup

#247 - Reduce the effect of common numbers twiddler

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: analysis, cleanup, humansolve

#246 - a-b-tester -h option doesn't print help and exit

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: bug, analysis

#245 - a-b-tester should only do one build of each tool, and should have a no-build option

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#244 - a-b-tester should print timing statistics (if -stats is provided)

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#243 - Consider making weka-trainer just use stdin/out, not files

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: analysis, cleanup

#242 - Consider a tool to automatically create different configs to a-b test

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#241 - Allow a-b-tester to stash and stash pop

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, analysis

#240 - Make it possible for a-b-tester to take a configurable sample_rate argument

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#239 - Make it possible for a-b-tester to generate relativedifficulties.csv too

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, analysis

#238 - Figure out a way for a-b-tester to not save so many files

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#237 - Make a-b-tester take an n argument for number of runs to do and average together r2

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#236 - HumanSolveOptions should have an argument for which Twiddlers to use

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, humansolve

#235 - dokugen-analysis should take a -no-cache option and reset-cache

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: enhancement, analysis

#234 - Write a script that uses CLI automation for weka to quickly determine if a change was good or not

Issue - State: closed - Opened by jkomoros over 8 years ago - 2 comments
Labels: enhancement, analysis, cleanup

#233 - Twiddler: Prefer cells in groups that are more filled

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, humansolve

#232 - Twiddler: Prefer numbers that are more represented in the grid

Issue - State: closed - Opened by jkomoros over 8 years ago
Labels: enhancement, humansolve

#231 - Experimentally set the twiddle weights and curves

Issue - State: open - Opened by jkomoros over 8 years ago - 2 comments
Labels: analysis, humansolve

#230 - Describe Marks and Excludes/Possibles differ/rationalize

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: sudoku

#229 - Twiddler: more votes for impossible, higher the probability

Issue - State: open - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, humansolve

#228 - A deep learned model of given a puzzle state where to look for moves next?

Issue - State: open - Opened by jkomoros over 8 years ago
Labels: enhancement, humansolve

#227 - Notion of refinding possibilities each technique

Issue - State: open - Opened by jkomoros over 8 years ago - 2 comments
Labels: enhancement, humansolve

#226 - Count human solve technique

Issue - State: open - Opened by jkomoros over 8 years ago - 2 comments
Labels: enhancement, humansolve

#225 - make i-sudoku store a log of all moves

Issue - State: open - Opened by jkomoros over 8 years ago - 5 comments
Labels: enhancement, analysis, humansolve, i-sudoku

#224 - Record the number/mark actions for real solves, score solves based on how close they are

Issue - State: open - Opened by jkomoros over 8 years ago - 1 comment
Labels: analysis, humansolve

#206 - Verify tweakChainedStepsWeights actually works

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: cleanup, humansolve

#201 - In some very weird cases in i-sudoku, the puzzle can show solved

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: bug, i-sudoku

#180 - Show difficulty of current puzzle

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, i-sudoku

#150 - Figure out if Weka can be controlled from the command line

Issue - State: closed - Opened by jkomoros over 8 years ago - 1 comment
Labels: enhancement, analysis

#143 - Make a makeReadOnly mode for grids

Issue - State: closed - Opened by jkomoros over 8 years ago - 12 comments
Labels: cleanup

#125 - Cull unnecessary non-fill steps in finished solvedirections

Issue - State: closed - Opened by jkomoros almost 9 years ago - 5 comments
Labels: enhancement, humansolve

#121 - Create a notion of "twiddlers" for step probabilities

Issue - State: closed - Opened by jkomoros almost 9 years ago - 1 comment
Labels: enhancement, humansolve

#48 - NonFill steps that don't end up being directly necessary to place a cell should be omitted

Issue - State: closed - Opened by jkomoros about 9 years ago - 3 comments
Labels: enhancement, humansolve

GitHub / jkomoros/sudoku issues and pull requests