Releases: sdfordham/pysyncon
Releases · sdfordham/pysyncon
v1.5.1
What's Changed
- Remove a coercion to integers in PlaceboTest code that can generate errors by @sdfordham in #63
- Add a check for duplicate rows in the foo argument of Dataprep by @sdfordham in #64
Full Changelog: v1.5.0...v1.5.1
v1.5.0
What's Changed
- A class
LinearFactorModel
was added that can generate data following a Linear factor model for testing/illustration/etc. purposes by @sdfordham in #51 - A
ConformalInference
class was added that can compute confidence intervals following Chernozhukov et al. A confidence interval method was added to theSynth
class that can generate confidence intervals using this class by @sdfordham in #56
Full Changelog: v1.4.0...v1.5.0
v1.4.0
What's Changed
- Added metric methods (APE, MAPE, MAE) to quantify the goodness of fit with the treated unit by @sdfordham in #45
- Now allow more statistical operations (sum, count, maximum, minimum, variance) on predictors (
predictors_op
argument ofDataprep
) by @sdfordham in #46
Full Changelog: v1.3.0...v1.4.0
Version 1.3.0
What's Changed
- The implementation of
PenalizedSynth
was flawed and has now been fixed so that it is faithful to the original paper (ref #31), by @sdfordham in #42
Full Changelog: v1.2.0...v1.3.0
Version 1.2.0
What's Changed
- Added
att
method to synthetic control methods that returns the average treatment effect on the treated unit over a given time period by @sdfordham in #32 - Dataprep string representation added by @sdfordham in #34
- Small speed ups in numpy array creations by @sdfordham in #36
Full Changelog: v1.1.0...v1.2.0
v1.1.0
v1.0.1
Release 1.0.0
- Add a method for running Robust Synthetic Control.
- Add a method for running Penalized Synthetic Control.
- Add a method for running Placebo Tests.
Release 0.2.1
- Fix missing legend on
path_plot
andgap_plot
Release 0.2.0
- Add basic augsynth functionality.
- Make compatible with python 3.8