v0.6.8

StatsModels v0.6.8

Diff since v0.6.7

Merged pull requests:

• Implement HypothesisCoding and un-export deprecate ContrastsCoding (#158) (@kleinschmidt)
• Install TagBot as a GitHub Action (#170) (@JuliaTagBot)

v0.6.7

v0.6.7 (2019-11-12)

Diff since v0.6.6

Closed issues:

• DataFrame not defined error when including interval keyword argument (#159)

Merged pull requests:

• add upper bounds to please the automerge bots (#166) (kleinschmidt)
• Fix typo in manual (#163) (nalimilan)
• get rid of CategoricalArrays (#157) (kleinschmidt)

v0.6.6

v0.6.6 (2019-10-10)

Diff since v0.6.5

Closed issues:

• one sided formulae no longer work (#143)

Merged pull requests:

• Fix predict with confidence interval (#160) (nalimilan)

v0.6.5

v0.6.5 (2019-10-02)

Diff since v0.6.4

Merged pull requests:

• Document support for special terms at run-time, one-sided formula (#155) (kleinschmidt)

v0.6.4

v0.6.4 (2019-09-29)

Diff since v0.6.3

Closed issues:

• Deprecated Missings.T (#150)
• DummyCoding and Missing (#148)
• Method ambiguity for lead/lag terms (#146)
• Stack overflow with views (#142)
• Interactions of the form x&id (#140)

Merged pull requests:

• bump patch version (#156) (kleinschmidt)
• Specialize apply_schema to Schema for lead/lag (#152) (kleinschmidt)
• Drop Missings and use Base.nonmissingtype (#151) (kleinschmidt)
• Fix intercept detector traits and hasresponse (#144) (kleinschmidt)
• Clean up tuples with sum after apply_schema (#127) (kleinschmidt)

v0.6.3

v0.6.3 (2019-08-02)

Diff since v0.6.2

Closed issues:

• Stack overflow for modelcols before apply\_schema (#136)
• Why is ModelMatrix assign = [1, 1, ..., 1] in v0.6.1? (#133)

Merged pull requests:

• Fix deprecated DataFrames syntax from tests (#138) (kleinschmidt)
• Make modelcols(::Term, d) an error (#137) (kleinschmidt)

v0.6.2

v0.6.2 (2019-07-18)

Diff since v0.6.1

Closed issues:

• Update StatsModels.jl to new DataFrames.jl 0.19 release (#132)
• predict\(model, df\) fails if df does not contain dependent variable (ERROR: type NamedTuple has no field) (#125)

Merged pull requests:

• Handle MatrixTerm when calculating column-to-term indexes in ModelMatrix (#134) (kleinschmidt)

v0.6.1

v0.6.1 (2019-06-29)

Diff since v0.6.0

Closed issues:

• Updating to the new version of StatsModels (#124)
• Release Terms 2.0 era for General (#100)

Merged pull requests:

• Fix bug with predict(mm::TableRegressionModel, data; kwargs...) (#126) (DilumAluthge)

v0.6.0

v0.6.0 (2019-06-25)

Diff since v0.5.0

Closed issues:

• Docs badge is broken (#102)
• Docs refer to nonexistent extract_matrix_term function (#101)
• Calling apply\_schema twice should not be an error (#97)
• Setting contrasts changes the number of columns in a model matrix (#95)
• Are offsets supported? (#92)
• With no intercept, number of model matrix columns depend on type of dependent variable (#48)
• default intercept behavior (#31)
• applying transformations on formula arguments (#25)
• An alternative representation of Terms? (#8)

Merged pull requests:

• add Compat bounds for Julia (#122) (oxinabox)
• Document to consider four times for internal transformation (#118) (oxinabox)
• Add Basic Lag and Lead terms (#108) (oxinabox)
• just copy continuous terms in modelcols, don't convert to float64 (#106) (kleinschmidt)
• next release will be version 0.6.0 (#105) (kleinschmidt)
• RIP REQUIRE, hello Project.toml (#104) (kleinschmidt)
• don't create schema for terms that already have them (#103) (kleinschmidt)
• Support keyword arguments in coeftable(::TableModels) (#96) (nalimilan)
• Bug-fix for respecting levels (#91) (Nosferican)
• mention that full-dummy is the same as one-hot in ML literature. (#89) (oxinabox)
• Fix deprecations (#87) (femtocleaner[bot])
• Stop using Compat since it's no longer needed (#82) (andreasnoack)
• Terms 2.0: son of Terms (#71) (kleinschmidt)

Version 0.5.0

v0.5.0

Make predict accept NamedTuple and return a DataFrame (#81)