Skip to content

Tags: dreamsxin/gota

Tags

v2.2.0

Toggle v2.2.0's commit message
v2.2.0: Correctness and Polish

v2.0.0

Toggle v2.0.0's commit message
v2.0.0: major feature release

DataFrame:
- Shift, Assign, Explode, Query (expr filter), Stack/Unstack, Resample
- CapplyParallel, RapplyParallel, AggregationParallel
- RenameAll, AsCategorical, WriteXLSXMultiSheet
- GroupBy.Transform row-order preservation, Time key support
- WithSheet option for ReadXLSX
- WriteSQL named placeholders (PostgreSQL/SQL Server)
- ScanCSV streaming mode, ReadNDJSON/WriteNDJSON

Series:
- Clip, Replace, Between, IsIn
- Abs, Round, Sign, Pow, Sqrt, Log, Log10, Exp
- Categorical type (dictionary encoding)
- Time type: Copy/Append/Subset/Fill fixed, comprehensive tests
- Rolling.StdDev Welford O(n), EWM.Var pandas-compatible formula
- FillNaNForwardLimit/BackwardLimit

Bug fixes:
- BatchConvert pool use-after-free
- Aggregation panic on empty groups
- Describe panic on Time columns
- Sample row-order preservation
- Query word-boundary operator matching
- GetGroups/Apply hidden column leak
- mat.Div division by zero returns NaN
- ScanCSV batch data integrity
- ReadNDJSON 10MB scanner buffer

v1.1.0

Toggle v1.1.0's commit message
feat: add new features with comprehensive tests and update README

Series enhancements:
- Add CumSum, CumProd, CumMax, CumMin cumulative statistics
- Add Diff and PctChange with configurable periods
- Add FillNaNForwardLimit and FillNaNBackwardLimit with fill count cap
- Add Corr and Cov (Pearson correlation and sample covariance)
- Add EWM (Exponentially Weighted Moving): Mean/Var/Std, span/alpha/adjust/minPeriods
- Fix EWM adjusted mode weight ordering to match pandas ewm(adjust=True)
- Extend Rolling window with Sum, Min, Max, Apply, MinPeriods

DataFrame enhancements:
- Add CumSum and CumProd with optional column subset
- Add Diff and PctChange with configurable periods and subset
- Add FillNAStrategyLimit (forward/backward fill with limit)
- Add Corr and Cov correlation/covariance matrix methods
- Add Melt for wide-to-long reshaping (pandas.melt equivalent)
- Add GroupBy.Apply and GroupBy.Transform
- Add hash-based join implementation (join_hash.go)

New I/O backends:
- Add ReadXLSX / WriteXLSX / ReadXLSXFile / WriteXLSXFile (excelize, no CGO)
- Add FromSQL (build DataFrame from *sql.Rows)
- Add WriteSQL with batch insert, CreateTable and TruncateFirst options

Index system:
- Add Index (single-level label index) with Get/Slice/Contains/IsUnique
- Add MultiIndex (hierarchical index) with partial key lookup
- Add IndexedDataFrame: WithIndex, WithColumnIndex, Loc, LocSlice, ResetIndex
- Add MultiIndexedDataFrame: WithMultiIndex, Loc (partial key)

Test coverage:
- series/series_extensions_test.go: 15 tests for new Series methods
- series/ewm_test.go: 8 tests for EWM Mean/Var/Std
- dataframe/dataframe_extensions_test.go: 15 tests for new DataFrame methods
- dataframe/excel_test.go: 3 round-trip tests for XLSX I/O
- dataframe/sql_test.go: 7 integration tests using modernc.org/sqlite (no CGO)
- dataframe/index_test.go: 22 tests for Index/MultiIndex/IndexedDataFrame

Documentation:
- Rewrite README with Table of Contents and full usage examples for all
  new features: EWM, cumulative stats, Diff/PctChange, FillNA with limit,
  Corr/Cov, Melt, GroupBy.Apply/Transform, Excel I/O, SQL I/O, Index/MultiIndex

v1.0.16

Toggle v1.0.16's commit message
Fix matrix Add

v1.0.15

Toggle v1.0.15's commit message
Fix matrix Add

v1.0.14

Toggle v1.0.14's commit message
Add dateframe::GetRow

v1.0.13

Toggle v1.0.13's commit message
Add TestLoadStructsWithSkipNames

v1.0.12

Toggle v1.0.12's commit message
Add dataframe::Show

v1.0.11

Toggle v1.0.11's commit message
Add series.Out

v1.0.10

Toggle v1.0.10's commit message
Add Series::Fill