* function code
* add tests for regressor coefficients utility
* add documentation for regressor_coefficients util function
* generate Rd docs
* add regressor_coefficients to R namespace
* minor formatting nit
* fix bugs
* added `flat_growth_init()` function
* added validation for 'flat'
* changed `fit.prophet()` to handle `growth='flat'`
* added `trend='flat'` capabilities to `sample_predictive_trend()` and `fit.prophet()`
* updated STAN code to handle flat trend
* [Syntax fix] Removed unnecessary bracket
* updated documentation
* undid formatting that was accidentally applied by autoformatter
* undid more formatting that was accidentally applied by autoformatter
* added tests
* typo in `sample_predictive_trend()`
* updated notebook with example in R
* updated documentation
* add test and initial function for mdape in R
* Add separate rolling_median_func and tests
* Modify rolling median function
* fix syntax in rolling median function
* sort by h
* R/diagnostics.R
* update .rd docs and notebook
* Add mdape to performance metrics params docstring
* Add test for custom cutoff cv
* implement custom cutoff logic in cv function
* add docstring
* add description in notebook and rebuild .Rd docs
* fix bug and add test case for period is NULL
* replace s.POSIXct set_date
I'm not sure why a data frame ends up with names and a tibble does not, but it doesn't seem like an important enough problem to investigate in detail, and this is a simple fix.
Ensures that `add_regressor` and `add_seasonality` are valid column names
to R, to ensure that the generated columns are then used downstream to fit
the model.
Why not put it in `validate_column_names`? Because `validate_column_names` is
also used to validate if holiday names (which can be scalar values in columns) are
valid. We want to allow `c('seans-bday', 'Xmas')` as a valid holiday input, so we
cannot then put it there.
Tested these changes by using devtools::testthat().
Resolves: #996
In reverse order of importance, this change addresses:
- #980, by adding validation to ensure that `fourier.order` > 0
- changes all the tests that `expect_error` using the argument
`fourier_order`, as the argument is named `fourier.order`, so
the intent of the test is to validate the holiday names rather
than to check the use of the correct argument name
Resolves: #980
* Allow both both hoidays and append holidays
Match holidays in predict and fit
Add test for append_holiday features; minor fixes
Add column name validation for append_holidays names; allow only one country
Fix bug about holiday year different from input year and also other bugs
Change function description
Add append holiday feature for R
* Add test for R/Pyrhon; fix bugs
* Add metrics for model validation
* Update to reflect comments on Github
* Change all_metrics() implementation to more sophisticated style
* Update metrics functions
* Remove redundant argument
* Use v0.3 files for diagnostics.R and prophet.R
* Use v0.3 file for man/predict.prophet.Rd
* Change the interface and do some refactoring
* Update memory requirement description per #326
* Fix R warning with extra regressor; disallow constant extra regressors.
* Fix unit test broken in new pandas
* Fix diagnostics unit tests for new pandas
* Fix copy with extra seasonalities / regressors Py
* Fix copy with extra seasonalities / regressors R
* Fix weekly_start and yearly_start in R plot_components
* Fix plotting in pandas 0.21 by using pydatetime instead of numpy
* Version bump
* Update README for new version
* Fix missing columns in SHF with extra regressor
* Update memory requirement description per #326
* Fix R warning with extra regressor; disallow constant extra regressors.
* Fix unit test broken in new pandas
* Fix diagnostics unit tests for new pandas
* Fix copy with extra seasonalities / regressors Py
* Fix copy with extra seasonalities / regressors R
* Fix weekly_start and yearly_start in R plot_components
* Fix plotting in pandas 0.21 by using pydatetime instead of numpy
