@@ -31,5 +31,5 @@ Suggests:
readr
License: BSD_3_clause + file LICENSE
LazyData: true
-RoxygenNote: 5.0.1
+RoxygenNote: 6.0.1
VignetteBuilder: knitr
@@ -514,6 +514,9 @@ fit.prophet <- function(m, df, ...) {
}
history <- df %>%
dplyr::filter(!is.na(y))
+ if (any(is.infinite(history$y))) {
+ stop("Found infinity in column y.")
+ }
m$history.dates <- sort(zoo::as.Date(df$ds))
out <- setup_dataframe(m, history, initialize_scales = TRUE)
@@ -16,4 +16,3 @@ Stan model.
\description{
Compile Stan model
-
@@ -14,4 +14,3 @@ df_for_plotting(m, fcst)
Merge history and forecast for plotting.
@@ -24,4 +24,3 @@ with the following elements:
sigma_obs (M array): Noise level.
Note that M=1 if MAP estimation.
@@ -19,4 +19,3 @@ Matrix with seasonality features.
Provides Fourier series components with the specified frequency and order.
@@ -15,4 +15,3 @@ array of indexes.
Gets changepoint matrix for history dataframe.
Load compiled Stan model
@@ -19,4 +19,3 @@ Provides a strong initialization for linear growth by calculating the
growth and offset parameters that pass the function through the first and
last points in the time series.
@@ -19,4 +19,3 @@ Provides a strong initialization for logistic growth by calculating the
@@ -17,4 +17,3 @@ Dataframe with seasonality.
Dataframe with seasonality features.
@@ -23,4 +23,3 @@ Dataframe that extends forward from the end of m$history for the
Make dataframe with future dates for forecasting.
@@ -17,4 +17,3 @@ A dataframe with a column for each holiday.
Construct a matrix of holiday features.
@@ -21,4 +21,3 @@ Dataframe with seasonality.
Data frame with seasonality features.
@@ -23,4 +23,3 @@ Vector y(t).
Evaluate the piecewise linear function.
@@ -25,4 +25,3 @@ Vector y(t).
Evaluate the piecewise logistic function.
@@ -41,4 +41,3 @@ plot(m, forecast)
@@ -19,4 +19,3 @@ A ggplot2 plot.
Plot the holidays component of the forecast.
@@ -20,4 +20,3 @@ A ggplot2 plot.
Plot the prophet trend.
@@ -21,4 +21,3 @@ A ggplot2 plot.
Plot the weekly component of the forecast.
Plot the yearly component of the forecast.
@@ -32,4 +32,3 @@ plot(m, forecast)
@@ -17,4 +17,3 @@ Dataframe with seasonal components.
Predict seasonality broken down into components.
@@ -17,4 +17,3 @@ Vector with trend on prediction dates.
Predict trend using the prophet model.
@@ -17,4 +17,3 @@ Dataframe with uncertainty intervals.
Prophet uncertainty intervals.
@@ -79,4 +79,3 @@ m <- prophet(history)
@@ -36,4 +36,3 @@ Plot the components of a prophet forecast.
Prints a ggplot2 with panels for trend, weekly and yearly seasonalities if
present, and holidays if present.
@@ -21,4 +21,3 @@ List of trend, seasonality, and yhat, each a vector like df$t.
Simulate observations from the extrapolated generative model.
@@ -19,4 +19,3 @@ Vector of simulated trend over df$t.
Simulate the trend using the extrapolated generative model.
@@ -17,4 +17,3 @@ Turns on yearly seasonality if there is >=2 years of history.
Turns on weekly seasonality if there is >=2 weeks of history, and the
spacing between dates in the history is <7 days.
@@ -20,4 +20,3 @@ Sets m$changepoints to the dates of changepoints. Either:
2) We are generating a grid of them.
3) The user prefers no changepoints be used.
@@ -21,4 +21,3 @@ Adds a time index and scales y. Creates auxillary columns 't', 't_ix',
'y_scaled', and 'cap_scaled'. These columns are used during both fitting
and predicting.
@@ -12,4 +12,3 @@ validate_inputs(m)
Validates the inputs to Prophet.
@@ -490,6 +490,8 @@ class Prophet(object):
raise Exception('Prophet object can only be fit once. '
'Instantiate a new object.')
history = df[df['y'].notnull()].copy()
+ if np.isinf(history['y'].values).any():
+ raise ValueError('Found infinity in column y.')
self.history_dates = pd.to_datetime(df['ds']).sort_values()
history = self.setup_dataframe(history, initialize_scales=True)