prophet-rb 0.3.2 → 0.4.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3e24dad0631694318db703e420f657e964911071b73e03cd73a6fc4c2d6e16fc
-  data.tar.gz: bcf2a53b3bcacf5461cd0b881bba0d026c7adee88200646f915074c9946711d3
+  metadata.gz: b7d18b97220bb11a0d836b1d272728ed9211aa85fb687032c6451d26b70c6755
+  data.tar.gz: c0de54e1a1f66785fe7a61f22c187376c43f9746f9cefa9f6d46834c742f15fb
 SHA512:
-  metadata.gz: bdad3069aeaadaa9ac8d339fb6e877942470fd5c8e11652bce50029b0a6de13d0fe383f6a97a343e29138396dec4b1c159e5ee87d229073d2f3eb5c084aa5af2
-  data.tar.gz: 1d54e502621ed4d34c818fe2e0e064578eae948957305300222892f4ee4750418e2471261596568d05795aea31bb4fe4639925da20a1f5db5bd20a0f0b35a66f
+  metadata.gz: 3b7d79d1b6c2fd8335ce417c653ecca5b7e15b9092dcba76f0062e38a775a0af6f88bb92334b2a24871745b2d64fb8822c3898617d9c391268cc7c7299e99e31
+  data.tar.gz: 0c094fe16a216ed0f19e566c5c5af896a878b3086f4ac0e550e58e9945c213aae9d12259a08c628b0f02662f7e7c25dcf10313ac65b311f9024a8b60f2b84331

data/CHANGELOG.md

@@ -1,3 +1,18 @@
+## 0.4.2 (2022-07-12)
+
+- Fixed warning with `add_country_holidays` method
+
+## 0.4.1 (2022-07-10)
+
+- Added support for cross validation and performance metrics
+- Added support for updating fitted models
+- Added support for saturating minimum forecasts
+
+## 0.4.0 (2022-07-07)
+
+- Added support for saving and loading models
+- Updated holidays
+
 ## 0.3.2 (2022-05-15)
 
 - Added advanced API options to `forecast` and `anomalies` methods

data/LICENSE.txt

@@ -1,7 +1,7 @@
 MIT License
 
 Copyright (c) Facebook, Inc. and its affiliates.
-Copyright (c) 2020 Andrew Kane
+Copyright (c) 2020-2022 Andrew Kane
 
 Permission is hereby granted, free of charge, to any person obtaining
 a copy of this software and associated documentation files (the

data/README.md

@@ -80,6 +80,18 @@ Prophet.anomalies(series, growth: "logistic", weekly_seasonality: false)
 
 Check out the [Prophet documentation](https://facebook.github.io/prophet/docs/quick_start.html) for a great explanation of all of the features. The advanced API follows the Python API and supports the same features. It uses [Rover](https://github.com/ankane/rover) for data frames.
 
+- [Quick Start](#advanced-quick-start)
+- [Plots](#plots)
+- [Saturating Forecasts](#saturating-forecasts)
+- [Trend Changepoints](#trend-changepoints)
+- [Holidays and Special Events](#holidays-and-special-events)
+- [Multiplicative Seasonality](#multiplicative-seasonality)
+- [Uncertainty Intervals](#uncertainty-intervals)
+- [Outliers](#outliers)
+- [Non-Daily Data](#non-daily-data)
+- [Diagnostics](#diagnostics)
+- [Additional Topics](#additional-topics)
+
 ## Advanced Quick Start
 
 [Explanation](https://facebook.github.io/prophet/docs/quick_start.html)
@@ -167,11 +179,24 @@ df = Rover.read_csv("example_wp_log_R.csv")
 df["cap"] = 8.5
 m = Prophet.new(growth: "logistic")
 m.fit(df)
-future = m.make_future_dataframe(periods: 365)
+future = m.make_future_dataframe(periods: 1826)
 future["cap"] = 8.5
 forecast = m.predict(future)
 ```
 
+Saturating minimum
+
+```ruby
+df["y"] = 10 - df["y"]
+df["cap"] = 6
+df["floor"] = 1.5
+future["cap"] = 6
+future["floor"] = 1.5
+m = Prophet.new(growth: "logistic")
+m.fit(df)
+forecast = m.predict(future)
+```
+
 ## Trend Changepoints
 
 [Explanation](https://facebook.github.io/prophet/docs/trend_changepoints.html)
@@ -204,11 +229,13 @@ Create a data frame with `holiday` and `ds` columns. Include all occurrences in
 ```ruby
 playoffs = Rover::DataFrame.new(
   "holiday" => "playoff",
-  "ds" => ["2008-01-13", "2009-01-03", "2010-01-16",
-           "2010-01-24", "2010-02-07", "2011-01-08",
-           "2013-01-12", "2014-01-12", "2014-01-19",
-           "2014-02-02", "2015-01-11", "2016-01-17",
-           "2016-01-24", "2016-02-07"],
+  "ds" => [
+    "2008-01-13", "2009-01-03", "2010-01-16",
+    "2010-01-24", "2010-02-07", "2011-01-08",
+    "2013-01-12", "2014-01-12", "2014-01-19",
+    "2014-02-02", "2015-01-11", "2016-01-17",
+    "2016-01-24", "2016-02-07"
+  ],
   "lower_window" => 0,
   "upper_window" => 1
 )
@@ -263,6 +290,8 @@ forecast = m.predict(future)
 
 [Explanation](https://facebook.github.io/prophet/docs/multiplicative_seasonality.html)
 
+Specify multiplicative seasonality
+
 ```ruby
 df = Rover.read_csv("example_air_passengers.csv")
 m = Prophet.new(seasonality_mode: "multiplicative")
@@ -271,8 +300,18 @@ future = m.make_future_dataframe(periods: 50, freq: "MS")
 forecast = m.predict(future)
 ```
 
+Specify mode when adding seasonality and regressors
+
+```ruby
+m = Prophet.new(seasonality_mode: "multiplicative")
+m.add_seasonality(name: "quarterly", period: 91.25, fourier_order: 8, mode: "additive")
+m.add_regressor("regressor", mode: "additive")
+```
+
 ## Uncertainty Intervals
 
+[Explanation](https://facebook.github.io/prophet/docs/uncertainty_intervals.html)
+
 Specify the width of uncertainty intervals (80% by default)
 
 ```ruby
@@ -285,6 +324,18 @@ Get uncertainty in seasonality
 Prophet.new(mcmc_samples: 300)
 ```
 
+## Outliers
+
+[Explanation](https://facebook.github.io/prophet/docs/outliers.html)
+
+Remove outliers
+
+```ruby
+df = Rover.read_csv("example_wp_log_R_outliers1.csv")
+df["y"][(df["ds"] > "2010-01-01") & (df["ds"] < "2011-01-01")] = Float::NAN
+m = Prophet.new.fit(df)
+```
+
 ## Non-Daily Data
 
 [Explanation](https://facebook.github.io/prophet/docs/non-daily_data.html)
@@ -298,6 +349,107 @@ future = m.make_future_dataframe(periods: 300, freq: "H")
 forecast = m.predict(future)
 ```
 
+## Diagnostics
+
+[Explanation](http://facebook.github.io/prophet/docs/diagnostics.html)
+
+Cross validation
+
+```ruby
+df_cv = Prophet::Diagnostics.cross_validation(m, initial: "730 days", period: "180 days", horizon: "365 days")
+```
+
+Custom cutoffs
+
+```ruby
+cutoffs = ["2013-02-15", "2013-08-15", "2014-02-15"].map { |v| Time.parse("#{v} 00:00:00 UTC") }
+df_cv2 = Prophet::Diagnostics.cross_validation(m, cutoffs: cutoffs, horizon: "365 days")
+```
+
+Get performance metrics
+
+```ruby
+df_p = Prophet::Diagnostics.performance_metrics(df_cv)
+```
+
+Plot cross validation metrics
+
+```ruby
+Prophet::Plot.plot_cross_validation_metric(df_cv, metric: "mape")
+```
+
+Hyperparameter tuning
+
+```ruby
+param_grid = {
+  changepoint_prior_scale: [0.001, 0.01, 0.1, 0.5],
+  seasonality_prior_scale: [0.01, 0.1, 1.0, 10.0]
+}
+
+# Generate all combinations of parameters
+all_params = param_grid.values[0].product(*param_grid.values[1..-1]).map { |v| param_grid.keys.zip(v).to_h }
+rmses = [] # Store the RMSEs for each params here
+
+# Use cross validation to evaluate all parameters
+all_params.each do |params|
+  m = Prophet.new(**params).fit(df) # Fit model with given params
+  df_cv = Prophet::Diagnostics.cross_validation(m, cutoffs: cutoffs, horizon: "30 days")
+  df_p = Prophet::Diagnostics.performance_metrics(df_cv, rolling_window: 1)
+  rmses << df_p["rmse"][0]
+end
+
+# Find the best parameters
+tuning_results = Rover::DataFrame.new(all_params)
+tuning_results["rmse"] = rmses
+p tuning_results
+```
+
+## Additional Topics
+
+[Explanation](https://facebook.github.io/prophet/docs/additional_topics.html)
+
+Save a model
+
+```ruby
+File.write("model.json", m.to_json)
+```
+
+Load a model
+
+```ruby
+m = Prophet.from_json(File.read("model.json"))
+```
+
+Uses the same format as Python, so models can be saved and loaded in either language
+
+Flat trend
+
+```ruby
+m = Prophet.new(growth: "flat")
+```
+
+Updating fitted models
+
+```ruby
+def stan_init(m)
+  res = {}
+  ["k", "m", "sigma_obs"].each do |pname|
+    res[pname] = m.params[pname][0, true][0]
+  end
+  ["delta", "beta"].each do |pname|
+    res[pname] = m.params[pname][0, true]
+  end
+  res
+end
+
+df = Rover.read_csv("example_wp_log_peyton_manning.csv")
+df1 = df[df["ds"] <= "2016-01-19"] # All data except the last day
+m1 = Prophet.new.fit(df1) # A model fit to all data except the last day
+
+m2 = Prophet.new.fit(df) # Adding the last day, fitting from scratch
+m2 = Prophet.new.fit(df, init: stan_init(m1)) # Adding the last day, warm-starting from m1
+```
+
 ## Resources
 
 - [Forecasting at Scale](https://peerj.com/preprints/3190.pdf)

data/data-raw/LICENSE-holidays.txt

@@ -0,0 +1,20 @@
+Copyright (c) 2017-2022 <maurizio.montel@gmail.com>
+Copyright (c) 2014-2017 <ryanssdev@icloud.com>
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

data/data-raw/README.md

@@ -0,0 +1,3 @@
+Holidays from 1995 through 2044 are generated from [this script](https://github.com/facebook/prophet/blob/main/python/scripts/generate_holidays_file.py).
+
+The data is primarily based on the Python [holidays](https://pypi.org/project/holidays/) package.