Generic equality function that raises an ``AssertionError`` if the objects are not equal. Notes ----- Works with pandas.DataFrame, pandas

Returns the short name.

Generates dataset for unit tests. :param freq: str pd.date_range freq parameter, e.g. H or D :param periods: int number of pe

Creates a forecast from input data and config. The result is also stored as ``self.forecast_result``. Parameters ----------

Adds a message to logger. Parameters ---------- message : `any` The message to be added to logger. level : `Enum` One

A function for forecasting. It captures growth, seasonality, holidays and other patterns. See "Capturing the time-dependence in the pr

Calculates the basic error measures in `~greykite.common.evaluation.EvaluationMetricEnum` and returns them in a dictionary. Parameters

Generates indices to split data into training and test CV folds according to rolling window time series cross validation Parameters

Generates dataset for unit tests that includes regressor columns :param freq: str pd.date_range freq parameter, e.g. H or D :param per

Plots the predicted anomalies over the actual anomalies. Parameters ---------- phase : str, default ``PHASE_PREDICT``

Returns the metric function.

Loads the Daily Peyton Manning dataset. This dataset contains log daily page views for the Wikipedia page for Peyton Manning. One of

Applies the default Forecast Config values to the given config. If an expected attribute value is provided, the value is unchanged. Ot

Finds trend change points automatically by adaptive lasso. The algorithm does an aggregation with a user-defined frequency, defaults daily.

Computation pipeline for end-to-end forecasting. Trains a forecast model end-to-end: 1. checks input data 2. runs cross-validati

Plots mean, quantiles, and overlays by the requested grouping dimension. The grouping dimension goes on the x-axis, and one line is shown for

Summarizes CV results for each grid search parameter combination. While ``grid_search.cv_results_`` could be imported into a `pandas.DataFram

Performs predictions using silverkite model. It determines if the prediction should be simulation-based or not and then predicts using

Adds default key-value pairs to items in ``overwrite_dict``. Merges the items in ``default_dict`` and ``overwrite_dict``, preferring ``overwr

Fits all transformers, transforms the data and concatenates results. Modified from `sklearn.pipeline.FeatureUnion`. Parameters

Helper function that validates forecast_pipeline output. Raises an AssertionError is results do not match the expected values. Parameters

Deserializes the output of the `dumps` method. Parameters ---------- serialized_dict: `dict` The output of the `d

Returns column name corresponding to a particular fourier term, as returned by fourier_series_fcn :param k: int fourier term :param c

Fits the quantile regression model. Parameters ---------- X : `numpy.array`, `pandas.DataFrame` or `pandas.Series`

Returns predictions on new data using the machine-learning (ml) model fitted via ``fit_ml_model``. :param fut_df: `pd.DataFrame` Inpu

Uses Depth First Search (DFS) to recursively serialize the input `obj`. For each object, the following happens: 1. If the `obj` is se

Loads data to internal representation. Parses date column, sets timezone aware index. Checks for irregularities and raises an error if input i

Loads data to internal representation. Parses date column, sets timezone aware index. Checks for irregularities and raises an error if

This function extracts a holiday data frame for the period of interest [year_start to year_end] for the given countries. This is done using th

Returns the greater_is_better boolean.

For a given function ``func`` which returns multiple outputs accessible by ``[]`` e.g. python list or dictionary, it construct a new function whic

A lightweight, easy-to-use function to create a plotly figure of given - lines (curves) - markers (points) - filled bands (e.g. error ban

Fits the lag based forecast model. Parameters ---------- X: `pandas.DataFrame` Input timeseries, with timestamp c

Returns the folder path ``data_dir/data_sub_dir``. If ``data_dir`` is None returns the internal data directory. By default the Greykit

Plots one or more lines against the same x-axis values. Parameters ---------- df : `pandas.DataFrame` Data frame with ``x_col`` a

This is the forecast function which can be used to forecast. It accepts extra regressors (``extra_pred_cols``) originally in ``df`` vi

Pre-processing before fitting ``Silverkite`` forecast model. Parameters ---------- X: `pandas.DataFrame` Input ti

Fits the predictive model and returns the prediction function. Parameters ---------- x_train : `numpy.array` The design matrix

Gets ``num_colors`` most distinguishable colors. Uses color maps "tab10", "tab20" or "viridis" depending on the number of colors needed. S

Keeps finite elements from reference_array, and corresponding elements in *arrays. Parameters ---------- reference_arrays : `list` [`nump

Sets default value for origin_for_time_vars Parameters ---------- df : `pandas.DataFrame` Training data. A data frame which inclu

Creates forecast for the dates specified in ``X``. Parameters ---------- X: `pandas.DataFrame` Input timeseries w

Default scorer for the estimator. Returns error based on ``score_func``. Because this is often used as null model, it is more informative to r

Returns a summary of the fitted model.

This function gets a datetime-like vector and creates new columns containing temporal features useful for time series analysis and forecasting e.g

Constants used by the template class. Includes the model templates and their default values.

Converts parameters of :func:`~greykite.algo.forecast.silverkite.forecast_simple_silverkite` into those of :func:`~greykite.algo.forec

Fits the ``transform_matrix`` based on input data, constraint, and objective function. Sets the attributes between ``forecasts`` and ``object

Plot the ``Prophet`` forecast components on the dataset passed to ``predict``. Will plot whichever are available of: trend, holidays,

Performs predictions for the dates in ``fut_df``. If ``extra_pred_cols`` refers to a column in ``df``, either ``fut_df`` or ``new_exte

Adds default key-value pairs to items in ``overwrite_dicts``. Merges the items in ``default_dict`` and ``overwrite_dicts``, preferring ``over

This function takes: - a time series, in the form of a dataframe: ``df`` - the anomaly information, in the form of a dataframe: ``ano

Uses DFS to recursively dump an object to pickle files. Originally intended for dumping the `~greykite.framework.pipeline.pipeline.ForecastRes

Flexible aggregation. Generates additional columns for evaluation via ``map_func_dict``, groups by ``groupby_col``, then aggregates according

Gets the changepoint dates from ``changepoints_dict`` Parameters ---------- changepoints_dict : `dict` or `None` The ``changepoin

Returns a ``pandas.DataFrame`` containing the dataset from ``data_path/data_name``. The input data must be in ``.csv`` or ``.csv.xz`` format.

Returns a ``pipeline`` argument to ``forecast_pipeline`` that uses ``DummyEstimator`` to make it easy to unit test ``forecast_pipeline``.

Computes mean, quantiles, and overlays by the requested grouping dimension. Overlays are best explained in the plotting context. The grouping

Create and returns custom pipeline parameters

Recursively removes dirs and files in ``dir_name``. This functions removes everything in ``dir_name`` that it has permission to remove. This

Adds a time feature data frame to a data frame by calling `~greykite.common.features.timeseries_features.build_time_features_df`. Parameters

Finds the seasonality change points (defined as the time points where seasonality magnitude changes, i.e., the time series becomes "fatter" or

Fits prediction models to continuous response vector (y) and report results. Parameters ---------- df : `pandas.DataFrame` A

Returns the number of splitting iterations yielded by the cross-validator Parameters ---------- X : array-like, shape (n_samp

Extends the input data for prediction into the future. Includes the historical values (VALUE_COL) so this can be fed into a Pipeline

Rounds numbers to significant figure. Rounds numbers to a specified significant figure, and uses string to format it. The number will be disp

Overrides the set_params method in `sklearn.base.BaseEstimator` to recognize estimator parameters. The set_params in sklearn requires

Extracts a column to group by from ``df``. Exactly one of ``groupby_time_feature``, ``groupby_sliding_window_size``, `groupby_custom_column`

Fits ``Silverkite`` forecast model. Parameters ---------- X: `pandas.DataFrame` Input timeseries, with timestamp

Returns integer value from input, with basic validation Parameters ---------- val : `float` or None, default None Value to conver

Get columns names from a list that matches ``pos_pattern``, but does not match ``neg_pattern``. If a column name matches both ``pos_pattern``

Loads the pickled files which are pickled by `~greykite.framework.templates.pickle_utils.dump_obj`. Originally intended for loading the `~

Converts date to continuous time. Each year is one unit. Parameters ---------- dt : datetime object the date to convert Retu

Makes important values of ``self.model_dict`` conveniently accessible. To be called by subclasses at the end of their ``fit`` method. Sets

Returns an uncertainty_dict for `~greykite.algo.forecast.silverkite.forecast_silverkite.SilverkiteForecast.forecast` input parameter: uncertai

Creates forecast for the dates specified in ``X``. Parameters ---------- X: `pandas.DataFrame` Input timeseries w

Checks the template name is valid and whether it is single or multi template. Raises an error if the template is not recognized. A va

A function to calculate confidence intervals (CI) for values given in ``distribution_col``. We allow for calculating as many quantiles as need

Fits ``Silverkite`` forecast model. Parameters ---------- X: `pandas.DataFrame` Input timeseries, with timestamp

Fits prophet model. Parameters ---------- X : `pandas.DataFrame` Input timeseries, with timestamp column,

Fits ``MultistageForecast`` forecast model. Parameters ---------- X: `pandas.DataFrame` Input timeseries, with ti

Generate data for testing the fitting algorithm with different parameters e.g. linear and random forest. :param n: integer sample

Returns a score function that pre-aggregates inputs according to ``agg_func``, and filters out invalid true values before evaluation. This allows

Infers significant holidays and holiday configurations. The class works for daily and sub-daily data. Sub-daily data is aggregated in

Loads the Hourly Bike Sharing Count dataset with possible aggregations. This dataset contains aggregated hourly count of the number of rented

This function takes a dataframe ``fut_df`` which includes the timestamps to forecast and a ``trained_model`` returned by `~greykite.al

Creates forecast for the dates specified in ``X``. Parameters ---------- X: `pandas.DataFrame` Input timeseries w

Returns an offset to add to test set indices when creating CV splits CV splits are shifted so that the last test observation is the last point

Performs the adaptive lasso cross-validation. Algorithm is based on a transformation of `sklearn.linear_model.LassoCV()`. If initial_coef i

Returns hyperparameter grid. Implements the method in `~greykite.framework.templates.base_template.BaseTemplate`. Uses ``self.time_p

Returns SimpleTimeFrequencyEnum based on input data period :param period: float Period of each observation (i.e. average time between obse

Returns the smallest timestamp that is greater than or equal to `ts` and is also a multiple of the `freq`. Assume hourly frequency i.e. `freq`

Returns the smallest gap between observations in df[time_col]. Assumes df[time_col] is sorted in ascending order without duplicates. :param

Class method to plot the components of a ``Silverkite`` model on datasets passed to either ``fit`` or ``predict``. Parameters