Version 0.14.0

• Use new Tables.jl interface.
• Improve printing.
• Improve documentation.
• Omit prefix in when using function name to create column names with aggregate.
• Fix readtable.
• Avoid copying a vector in stackdf.
• Raise errors when passing a single Boolean to getindex.
• Fix stack overflow with vcat and by/aggregate.
• Deprecate iterating over DataFrameRow in favor of pairs.
• Add DataFrame constructor accepting a DataFrame.

Version 0.13.1

Merge pull request #1487 from ExpandingMan/fix-int-bit-assert

fixed overzealous type assertion

Version 0.13.0

Merge pull request #1471 from JuliaData/jq/1.0

Updates to run on 1.0

v0.12.0

Fix a deprecation warning (#1463)

Version 0.11.7

v0.11.7

Fix printing of nothings (#1454)

Version 0.11.6

v0.11.6

Fix push!(df, a, b, ...) (#1398)

Version 0.11.5

• Fix combining DataFrames with a column with element type Missing.
• Test non-matching joins.

Version 0.11.4

• Add filter and filter! methods.
• Deprecate automatic deduplication of column names, add makeunique keyword argument to enable it.
• Improve DataFrame constructors and conversions for Vector and Matrix.
• Check categorical argument length in constructor.
• Redesign getindex and fix stack/melt StackOverflowError.
• Fix a stack method calling stackdf instead of stack.
• Fix readtable with zipped files.
• Add deprecation for vcat(Vector{<:AbstractDataFrame}).

Version 0.11.3

• Fix similar and allowmissing!.
• Support joining dataframes on columns with different left/right names.
• Improve unstack.
• Keep columns which do not allow for missing values as such in join.
• Fix hash when passed an initial hash value.
• Make boolean indexing stricter to match behavior used in Julia Base.
• Julia 0.7 fixes.

Version 0.11.2

• Fix readtable in the presence of missing values.
• Use CodecZlib instead of GZip for readtable.
• Add one-argument version of allowmissing! to change all columns to accept missing values.
• Deprecate pipelining with |> and currying functions.
• Correct docstring for combine.
• Fix unstack in some special cases.
• Fix hcat between a vector as first argument and a DataFrame as second argument.
• Fix Julia 0.7 deprecations.