saymrwulf/prophet
1
0
Fork 0
mirror of https://github.com/saymrwulf/prophet.git synced 2026-09-16 22:20:23 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Documentation for cross validation

Browse source
This commit is contained in:
bl 2017-09-02 10:53:38 -07:00
parent 6c446cee85
commit 2f9b20b2d3
6 changed files with 332 additions and 81 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

51
R/R/diagnostics.R
View file

@ -27,19 +27,21 @@ generate_cutoffs <- function(df, horizon, k, period) {
} }
tzone <- attr(cutoff, "tzone") # Timezone is wiped by putting in array tzone <- attr(cutoff, "tzone") # Timezone is wiped by putting in array
result <- c(cutoff) result <- c(cutoff)
for (i in 2:k) { if (k > 1) {
cutoff <- cutoff - period for (i in 2:k) {
# If data does not exist in data range (cutoff, cutoff + horizon] cutoff <- cutoff - period
if (!any((df$ds > cutoff) & (df$ds <= cutoff + horizon))) { # If data does not exist in data range (cutoff, cutoff + horizon]
# Next cutoff point is 'closest date before cutoff in data - horizon' if (!any((df$ds > cutoff) & (df$ds <= cutoff + horizon))) {
closest.date <- max(df$ds[df$ds <= cutoff]) # Next cutoff point is 'closest date before cutoff in data - horizon'
cutoff <- closest.date - horizon closest.date <- max(df$ds[df$ds <= cutoff])
cutoff <- closest.date - horizon
}
if (cutoff < min(df$ds)) {
warning('Not enough data for requested number of cutoffs! Using ', i)
break
}
result <- c(result, cutoff)
} }
if (cutoff < min(df$ds)) {
warning('Not enough data for requested number of cutoffs! Using ', i)
break
}
result <- c(result, cutoff)
} }
# Reset timezones # Reset timezones
attr(result, "tzone") <- tzone attr(result, "tzone") <- tzone
@ -47,8 +49,9 @@ generate_cutoffs <- function(df, horizon, k, period) {
} }
#' Simulated historical forecasts. #' Simulated historical forecasts.
#' Make forecasts from k historical cutoff dates, and compare forecast values #'
#' to actual values. #' Make forecasts from k historical cutoff points, working backwards from
#' (end - horizon) with a spacing of period between each cutoff.
#' #'
#' @param model Fitted Prophet model. #' @param model Fitted Prophet model.
#' @param horizon Integer size of the horizon #' @param horizon Integer size of the horizon
@ -99,25 +102,31 @@ simulated_historical_forecasts <- function(model, horizon, units, k,
} }
#' Cross-validation for time series. #' Cross-validation for time series.
#' Computes forecast error with cutoffs at the specified period. When the #'
#' period is the time interval of the data, is the procedure described in #' Computes forecasts from historical cutoff points. Beginning from initial,
#' https://robjhyndman.com/hyndsight/tscv/. Beginning from end-horizon, makes #' makes cutoffs with a spacing of period up to (end - horizon).
#' a cutoff every "period" amount of time, going back to "initial". #'
#' When period is equal to the time interval of the data, this is the
#' technique described in https://robjhyndman.com/hyndsight/tscv/ .
#' #'
#' @param model Fitted Prophet model. #' @param model Fitted Prophet model.
#' @param horizon Integer size of the horizon #' @param horizon Integer size of the horizon
#' @param units String unit of the horizon, e.g., "days", "secs". #' @param units String unit of the horizon, e.g., "days", "secs".
#' @param period Integer amount of time between cutoff dates. Same units as #' @param period Integer amount of time between cutoff dates. Same units as
#' horizon. #' horizon. If not provided, 0.5 * horizon is used.
#' @param initial Integer size of the first training period. If not provided, #' @param initial Integer size of the first training period. If not provided,
#' 3 * horizon is used. Same units as horizon. #' 3 * horizon is used. Same units as horizon.
#' #'
#' @return A dataframe with the forecast, actual value, and cutoff date. #' @return A dataframe with the forecast, actual value, and cutoff date.
#' #'
#' @export #' @export
cross_validation <- function(model, horizon, units, period, initial = NULL) { cross_validation <- function(
model, horizon, units, period = NULL, initial = NULL) {
te <- max(model$history$ds) te <- max(model$history$ds)
ts <- min(model$history$ds) ts <- min(model$history$ds)
if (is.null(period)) {
period <- 0.5 * horizon
}
if (is.null(initial)) { if (is.null(initial)) {
initial <- 3 * horizon initial <- 3 * horizon
} }
@ -129,7 +138,7 @@ cross_validation <- function(model, horizon, units, period, initial = NULL) {
as.double(period.dt, units = 'secs') as.double(period.dt, units = 'secs')
) )
if (k < 1) { if (k < 1) {
stop('Not enough data for specified horizon and initial.') stop('Not enough data for specified horizon, period, and initial.')
} }
return(simulated_historical_forecasts(model, horizon, units, k, period)) return(simulated_historical_forecasts(model, horizon, units, k, period))
} }

21
R/man/cross_validation.Rd
View file

@ -2,13 +2,9 @@
% Please edit documentation in R/diagnostics.R % Please edit documentation in R/diagnostics.R
\name{cross_validation} \name{cross_validation}
\alias{cross_validation} \alias{cross_validation}
\title{Cross-validation for time series. \title{Cross-validation for time series.}
Computes forecast error with cutoffs at the specified period. When the
period is the time interval of the data, is the procedure described in
https://robjhyndman.com/hyndsight/tscv/. Beginning from end-horizon, makes
a cutoff every "period" amount of time, going back to "initial".}
\usage{ \usage{
cross_validation(model, horizon, units, period, initial = NULL) cross_validation(model, horizon, units, period = NULL, initial = NULL)
} }
\arguments{ \arguments{
\item{model}{Fitted Prophet model.} \item{model}{Fitted Prophet model.}
@ -18,7 +14,7 @@ cross_validation(model, horizon, units, period, initial = NULL)
\item{units}{String unit of the horizon, e.g., "days", "secs".} \item{units}{String unit of the horizon, e.g., "days", "secs".}
\item{period}{Integer amount of time between cutoff dates. Same units as \item{period}{Integer amount of time between cutoff dates. Same units as
horizon.} horizon. If not provided, 0.5 * horizon is used.}
\item{initial}{Integer size of the first training period. If not provided, \item{initial}{Integer size of the first training period. If not provided,
3 * horizon is used. Same units as horizon.} 3 * horizon is used. Same units as horizon.}
@ -27,9 +23,10 @@ horizon.}
A dataframe with the forecast, actual value, and cutoff date. A dataframe with the forecast, actual value, and cutoff date.
} }
\description{ \description{
Cross-validation for time series. Computes forecasts from historical cutoff points. Beginning from initial,
Computes forecast error with cutoffs at the specified period. When the makes cutoffs with a spacing of period up to (end - horizon).
period is the time interval of the data, is the procedure described in }
https://robjhyndman.com/hyndsight/tscv/. Beginning from end-horizon, makes \details{
a cutoff every "period" amount of time, going back to "initial". When period is equal to the time interval of the data, this is the
technique described in https://robjhyndman.com/hyndsight/tscv/ .
} }

18
R/man/parse_seasonality_args.Rd
View file

@ -2,11 +2,8 @@
% Please edit documentation in R/prophet.R % Please edit documentation in R/prophet.R
\name{parse_seasonality_args} \name{parse_seasonality_args}
\alias{parse_seasonality_args} \alias{parse_seasonality_args}
\alias{parse_seasonality_args}
\title{Get number of Fourier components for built-in seasonalities.} \title{Get number of Fourier components for built-in seasonalities.}
\usage{ \usage{
parse_seasonality_args(m, name, arg, auto.disable, default.order)
parse_seasonality_args(m, name, arg, auto.disable, default.order) parse_seasonality_args(m, name, arg, auto.disable, default.order)
} }
\arguments{ \arguments{
@ -19,27 +16,12 @@ provided.}
\item{auto.disable}{Bool if seasonality should be disabled when 'auto'.} \item{auto.disable}{Bool if seasonality should be disabled when 'auto'.}
\item{default.order}{Int default Fourier order.}
\item{m}{Prophet object.}
\item{name}{String name of the seasonality component.}
\item{arg}{'auto', TRUE, FALSE, or number of Fourier components as
provided.}
\item{auto.disable}{Bool if seasonality should be disabled when 'auto'.}
\item{default.order}{Int default Fourier order.} \item{default.order}{Int default Fourier order.}
} }
\value{ \value{
Number of Fourier components, or 0 for disabled.
Number of Fourier components, or 0 for disabled. Number of Fourier components, or 0 for disabled.
} }
\description{ \description{
Get number of Fourier components for built-in seasonalities.
Get number of Fourier components for built-in seasonalities. Get number of Fourier components for built-in seasonalities.
} }
\keyword{internal} \keyword{internal}

9
R/man/simulated_historical_forecasts.Rd
View file

@ -2,9 +2,7 @@
% Please edit documentation in R/diagnostics.R % Please edit documentation in R/diagnostics.R
\name{simulated_historical_forecasts} \name{simulated_historical_forecasts}
\alias{simulated_historical_forecasts} \alias{simulated_historical_forecasts}
\title{Simulated historical forecasts. \title{Simulated historical forecasts.}
Make forecasts from k historical cutoff dates, and compare forecast values
to actual values.}
\usage{ \usage{
simulated_historical_forecasts(model, horizon, units, k, period = NULL) simulated_historical_forecasts(model, horizon, units, k, period = NULL)
} }
@ -24,7 +22,6 @@ horizon. If not provided, will use 0.5 * horizon.}
A dataframe with the forecast, actual value, and cutoff date. A dataframe with the forecast, actual value, and cutoff date.
} }
\description{ \description{
Simulated historical forecasts. Make forecasts from k historical cutoff points, working backwards from
Make forecasts from k historical cutoff dates, and compare forecast values (end - horizon) with a spacing of period between each cutoff.
to actual values.
} }

260
notebooks/diagnostics.ipynb Normal file
View file

File diff suppressed because one or more lines are too long

54
python/fbprophet/diagnostics.py
View file

@ -46,11 +46,13 @@ def _cutoffs(df, horizon, k, period):
cutoff -= period cutoff -= period
# If data does not exist in data range (cutoff, cutoff + horizon] # If data does not exist in data range (cutoff, cutoff + horizon]
if not (((df['ds'] > cutoff) & (df['ds'] <= cutoff + horizon)).any()): if not (((df['ds'] > cutoff) & (df['ds'] <= cutoff + horizon)).any()):
# Next cutoff point is 'closest date before cutoff in data - horizon' # Next cutoff point is 'last date before cutoff in data - horizon'
closest_date = df[df['ds'] <= cutoff].max()['ds'] closest_date = df[df['ds'] <= cutoff].max()['ds']
cutoff = closest_date - horizon cutoff = closest_date - horizon
if cutoff < df['ds'].min(): if cutoff < df['ds'].min():
logger.warning('Not enough data for requested number of cutoffs! Using {}.'.format(i)) logger.warning(
'Not enough data for requested number of cutoffs! '
'Using {}.'.format(i))
break break
result.append(cutoff) result.append(cutoff)
@ -60,20 +62,20 @@ def _cutoffs(df, horizon, k, period):
def simulated_historical_forecasts(model, horizon, k, period=None): def simulated_historical_forecasts(model, horizon, k, period=None):
"""Simulated Historical Forecasts. """Simulated Historical Forecasts.
If you would like to know it in detail, read the original paper
https://facebookincubator.github.io/prophet/static/prophet_paper_20170113.pdf Make forecasts from k historical cutoff points, working backwards from
(end - horizon) with a spacing of period between each cutoff.
Parameters Parameters
---------- ----------
model: Prophet class object. model: Prophet class object.
Fitted Prophet model Fitted Prophet model
horizon: string which has pd.Timedelta compatible style. horizon: string with pd.Timedelta compatible style, e.g., '5 days',
Forecast horizon ('5 days', '3 hours', '10 seconds' etc) '3 hours', '10 seconds'.
k: Int number. k: Int number of forecasts point.
The number of forecasts point. period: Optional string with pd.Timedelta compatible style. Simulated
period: string which has pd.Timedelta compatible style or None, default None. forecast will be done at every this period. If not provided,
Simulated Forecast will be done at every this period. 0.5 * horizon is used.
0.5 * horizon is used when it is None.
Returns Returns
------- -------
@ -108,21 +110,24 @@ def simulated_historical_forecasts(model, horizon, k, period=None):
return reduce(lambda x, y: x.append(y), predicts).reset_index(drop=True) return reduce(lambda x, y: x.append(y), predicts).reset_index(drop=True)
def cross_validation(model, horizon, period, initial=None): def cross_validation(model, horizon, period=None, initial=None):
"""Cross-Validation for time-series. """Cross-Validation for time series.
This function is the same with Time series cross-validation described in https://robjhyndman.com/hyndsight/tscv/
when the value of period is equal to the time interval of data. Computes forecasts from historical cutoff points. Beginning from initial,
makes cutoffs with a spacing of period up to (end - horizon).
When period is equal to the time interval of the data, this is the
technique described in https://robjhyndman.com/hyndsight/tscv/ .
Parameters Parameters
---------- ----------
model: Prophet class object. Fitted Prophet model model: Prophet class object. Fitted Prophet model
horizon: string which has pd.Timedelta compatible style. horizon: string with pd.Timedelta compatible style, e.g., '5 days',
Forecast horizon ('5 days', '3 hours', '10 seconds' etc) '3 hours', '10 seconds'.
period: string which has pd.Timedelta compatible style. period: string with pd.Timedelta compatible style. Simulated forecast will
Simulated Forecast will be done at every this period. be done at every this period. If not provided, 0.5 * horizon is used.
initial: string which has pd.Timedelta compatible style or None, default None. initial: string with pd.Timedelta compatible style. The first training
First training period. period will begin here. If not provided, 3 * horizon is used.
3 * horizon is used when it is None.
Returns Returns
------- -------
@ -131,9 +136,10 @@ def cross_validation(model, horizon, period, initial=None):
te = model.history['ds'].max() te = model.history['ds'].max()
ts = model.history['ds'].min() ts = model.history['ds'].min()
horizon = pd.Timedelta(horizon) horizon = pd.Timedelta(horizon)
period = pd.Timedelta(period) period = 0.5 * horizon if period is None else pd.Timedelta(period)
initial = 3 * horizon if initial is None else pd.Timedelta(initial) initial = 3 * horizon if initial is None else pd.Timedelta(initial)
k = int(np.ceil(((te - horizon) - (ts + initial)) / period)) k = int(np.ceil(((te - horizon) - (ts + initial)) / period))
if k < 1: if k < 1:
raise ValueError('Not enough data for specified horizon and initial.') raise ValueError(
'Not enough data for specified horizon, period, and initial.')
return simulated_historical_forecasts(model, horizon, k, period) return simulated_historical_forecasts(model, horizon, k, period)