diff --git a/DESCRIPTION b/DESCRIPTION
index cacd43ba..533d08b1 100755
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,4 +1,4 @@
-Package: epitools
+Package: epiprocess
Type: Package
Title: Tools for basic signal processing in epidemiology
Version: 1.0.0
diff --git a/NAMESPACE b/NAMESPACE
index 6ee6fd35..7e588c50 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -1,13 +1,13 @@
# Generated by roxygen2: do not edit by hand
-S3method(as.epi_tibble,data.frame)
-S3method(as.epi_tibble,epi_tibble)
-S3method(as.epi_tibble,tibble)
-S3method(group_by,epi_tibble)
-S3method(head,epi_tibble)
-S3method(print,epi_tibble)
-S3method(summary,epi_tibble)
-S3method(ungroup,epi_tibble)
+S3method(as.epi_df,data.frame)
+S3method(as.epi_df,epi_df)
+S3method(as.epi_df,tibble)
+S3method(group_by,epi_df)
+S3method(head,epi_df)
+S3method(print,epi_df)
+S3method(summary,epi_df)
+S3method(ungroup,epi_df)
export("%>%")
export(End)
export(Max)
@@ -16,11 +16,11 @@ export(Median)
export(Min)
export(Start)
export(Sum)
-export(as.epi_tibble)
-export(cor_lagged)
+export(as.epi_df)
export(detect_outliers)
export(detect_outliers_rm)
export(detect_outliers_stl)
+export(epi_cor)
export(epi_slide)
export(estimate_deriv)
export(pct_change)
diff --git a/R/correlation.R b/R/correlation.R
index d617bc3d..28c1abd1 100644
--- a/R/correlation.R
+++ b/R/correlation.R
@@ -1,12 +1,12 @@
-#' Compute lagged correlations between variables in an `epi_tibble` object
+#' Compute lagged correlations between variables in an `epi_df` object
#'
-#' Computes lagged correlations between variables in an `epi_tibble` object,
+#' Computes lagged correlations between variables in an `epi_df` object,
#' allowing for grouping by geo value, time value, or any other variables. See
-#' the [correlations
-#' vignette](https://cmu-delphi.github.io/epitools/articles/correlations.html)
+#' the [correlation
+#' vignette](https://cmu-delphi.github.io/epitools/articles/correlation.html)
#' for examples.
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var1,var2 The variables in `x` to correlate.
#' @param dt1,dt2 Time shifts to consider for the two variables, respectively,
#' before computing correlations. Negative shifts translate into in a lag
@@ -40,11 +40,11 @@
#' @importFrom stats cor
#' @importFrom rlang .data enquo
#' @export
-cor_lagged = function(x, var1, var2, dt1 = 0, dt2 = 0,
- by = geo_value, use = "na.or.complete",
- method = c("pearson", "kendall", "spearman")) {
- # Check we have an `epi_tibble` object
- if (!inherits(x, "epi_tibble")) abort("`x` must be of class `epi_tibble`.")
+epi_cor = function(x, var1, var2, dt1 = 0, dt2 = 0, by = geo_value,
+ use = "na.or.complete",
+ method = c("pearson", "kendall", "spearman")) {
+ # Check we have an `epi_df` object
+ if (!inherits(x, "epi_df")) abort("`x` must be of class `epi_df`.")
# Check that we have variables to do computations on
if (missing(var1)) abort("`var1` must be specified.")
diff --git a/R/derivative.R b/R/derivative.R
index 9453c26f..0061b386 100644
--- a/R/derivative.R
+++ b/R/derivative.R
@@ -1,12 +1,12 @@
-#' Estimate derivatives of a variable in an `epi_tibble` object
+#' Estimate derivatives of a variable in an `epi_df` object
#'
-#' Estimates derivatives of a variable in an `epi_tibble` object, using a local
-#' (in time) linear regression or alternative smoother. See the [estimating
-#' derivatives
-#' vignette](https://cmu-delphi.github.io/epitools/articles/derivatives.html)
-#' for examples.
+#' Estimates derivatives of a variable in an `epi_df` object, using a local (in
+#' time) linear regression or alternative smoothing method. See the [estimating
+#' derivatives
+#' vignette](https://cmu-delphi.github.io/epitools/articles/derivative.html) for
+#' examples.
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var The variable in `x` whose derivatives are to be estimated.
#' @param method One of "lin", "ss", or "tf" indicating the method to use for
#' the derivative calculation. To estimate the derivative at any time point,
@@ -49,7 +49,7 @@
#' is `TRUE`.
#' @param ... Additional arguments to pass to the function that estimates
#' derivatives. See details below.
-#' @return An `epi_tibble` object given by appending a new column to `x`, named
+#' @return An `epi_df` object given by appending a new column to `x`, named
#' according to the `new_col_name` argument, containing the derivative values.
#'
#' @details Derivatives are estimated using:
@@ -141,7 +141,7 @@ estimate_deriv = function(x, var, method = c("lin", "ss", "tf"), n = 14,
x = select(x, -temp) %>% ungroup()
# Attach the class and metadata and return
- class(x) = c("epi_tibble", class(x))
+ class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
return(x)
}
diff --git a/R/epi_tibble.R b/R/epi_df.R
similarity index 76%
rename from R/epi_tibble.R
rename to R/epi_df.R
index 8bc97c31..846cb918 100644
--- a/R/epi_tibble.R
+++ b/R/epi_df.R
@@ -1,8 +1,8 @@
-#' Convert data to `epi_tibble` format
+#' Convert data to `epi_df` format
#'
#' Converts a data frame or tibble into a format that is consistent with the
-#' `epi_tibble` class, ensuring that it has a certain minimal set of columns,
-#' and that it has certain minimal metadata.
+#' `epi_df` class, ensuring that it has a certain minimal set of columns, and
+#' that it has certain minimal metadata.
#'
#' @param x The object to be converted. See the methods section below for
#' details on formatting of each input type.
@@ -16,20 +16,19 @@
#' attempt to infer it from the passed object `x`; if this fails, then the
#' current day-time will be used.
#' @param additional_metadata List of additional metadata to attach to the
-#' `epi_tibble` object. All objects will have `time_type`, `geo_type`, and
-#' `issue` fields; named entries from the passed list or will be included as
-#' well.
+#' `epi_df` object. All objects will have `time_type`, `geo_type`, and `issue`
+#' fields; named entries from the passed list or will be included as well.
#' @param ... Additional arguments passed to methods.
-#' @return An `epi_tibble` object.
+#' @return An `epi_df` object.
#'
-#' @details An `epi_tibble` is a tibble with (at least) the following columns:
+#' @details An `epi_df` is a tibble with (at least) the following columns:
#'
#' * `geo_value`: the geographic value associated with each measurement.
#' * `time_value`: the time value associated with each measurement.
#'
#' Other columns can be considered as measured variables, which we also broadly
-#' refer to as signal variables. An `epi_tibble` object also has metadata with
-#' (at least) the following fields:
+#' refer to as signal variables. An `epi_df` object also has metadata with (at
+#' least) the following fields:
#'
#' * `geo_type`: the type for the geo values.
#' * `time_type`: the type for the time values.
@@ -37,19 +36,19 @@
#'
#' The first two fields above, `geo_type` and `time_type`, can usually be
#' inferred from the `geo_value` and `time_value` columns, respectively. The
-#' last field above, `issue`, is the most unique to the `epi_tibble` format.
-#' In a typical case, this represents the maximum of the issues of individual
+#' last field above, `issue`, is the most unique to the `epi_df` format. In a
+#' typical case, this represents the maximum of the issues of individual
#' signal values measured in the data set; hence we would also say that the
#' data set is comprised of all signal values observed "as of" the given issue
#' in the metadata.
#'
-#' Metadata for an `epi_tibble` object `x` can be accessed (and altered) via
+#' Metadata for an `epi_df` object `x` can be accessed (and altered) via
#' `attributes(x)$metadata`. More information on geo types, time types, and
#' issues is given below.
#'
#' @section Geo types:
-#' The following geo types are supported in an `epi_tibble`. Their geo coding
-#' (specification of geo values for each geo type) is also described below.
+#' The following geo types are supported in an `epi_df`. Their geo coding
+#' (specification of geo values for each geo type) is also described below.
#'
#' * `"county"`: each observation corresponds to a U.S. county; coded by 5-digit
#' FIPS code.
@@ -69,7 +68,7 @@
#' geo type is labeled as "custom".
#'
#' @section Time types:
-#' The following time types are supported in an `epi_tibble`. Their time coding
+#' The following time types are supported in an `epi_df`. Their time coding
#' (specification of time values for each time type) is also described below.
#'
#' * `"day-time"`: each observation corresponds to a time on a given day (measured
@@ -88,19 +87,19 @@
#' todo
#'
#' @export
-as.epi_tibble = function(x, ...) {
- UseMethod("as.epi_tibble")
+as.epi_df = function(x, ...) {
+ UseMethod("as.epi_df")
}
-#' @method as.epi_tibble epi_tibble
-#' @describeIn as.epi_tibble Simply returns the `epi_tibble` object unchanged.
+#' @method as.epi_df epi_df
+#' @describeIn as.epi_df Simply returns the `epi_df` object unchanged.
#' @export
-as.epi_tibble.epi_tibble = function(x, ...) {
+as.epi_df.epi_df = function(x, ...) {
return(x)
}
-#' @method as.epi_tibble tibble
-#' @describeIn as.epi_tibble The input tibble `x` must contain the columns
+#' @method as.epi_df tibble
+#' @describeIn as.epi_df The input tibble `x` must contain the columns
#' `geo_value` and `time_value`. All other columns will be preserved as is,
#' and treated as measured variables. If `issue` is missing, then the function
#' will look for `issue` as a column of `x`, or as a field in its metadata
@@ -108,8 +107,8 @@ as.epi_tibble.epi_tibble = function(x, ...) {
#' current day-time will be used.
#' @importFrom rlang .data abort
#' @export
-as.epi_tibble.tibble = function(x, geo_type, time_type, issue,
- additional_metadata = list(), ...) {
+as.epi_df.tibble = function(x, geo_type, time_type, issue,
+ additional_metadata = list(), ...) {
# Check that we have geo_value and time_value columns
if (!("geo_value" %in% names(x))) {
abort("`x` must contain a `geo_value` column.")
@@ -198,9 +197,9 @@ as.epi_tibble.tibble = function(x, geo_type, time_type, issue,
metadata$issue = issue
metadata = c(metadata, additional_metadata)
- # Convert to a tibble, apply epi_tibble class, attach metadata
+ # Convert to a tibble, apply epi_df class, attach metadata
if (!inherits(x, "tibble")) x = tibble::as_tibble(x)
- class(x) = c("epi_tibble", class(x))
+ class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
# Reorder columns (geo_value, time_value, ...) and return
@@ -208,30 +207,30 @@ as.epi_tibble.tibble = function(x, geo_type, time_type, issue,
return(x)
}
-#' @method as.epi_tibble data.frame
-#' @describeIn as.epi_tibble The input data frame `x` must contain the columns
+#' @method as.epi_df data.frame
+#' @describeIn as.epi_df The input data frame `x` must contain the columns
#' `geo_value` and `time_value`. All other columns will be preserved as is,
#' and treated as measured variables. If `issue` is missing, then the function
#' will look for `issue` as a column of `x`, or as a field in its metadata
#' (stored in its attributes), to infer the issue; if this fails, then the
-#' current day-time will be used.
+#' current day-time will be used.
#' @export
-as.epi_tibble.data.frame = as.epi_tibble.tibble
+as.epi_df.data.frame = as.epi_df.tibble
-#' Print `epi_tibble` object
+#' Print `epi_df` object
#'
-#' Prints a brief summary of the `epi_tibble` object, then prints the underlying
-#' tibble.
+#' Prints a brief summary of the `epi_df` object, then prints the underlying
+#' tibble.
#'
-#' @param x The `epi_tibble` object.
+#' @param x The `epi_df` object.
#' @param ... Additional arguments passed to `print.tibble()` to print the
#' data.
-#' @return The `epi_tibble` object, unchanged.
+#' @return The `epi_df` object, unchanged.
#'
-#' @method print epi_tibble
+#' @method print epi_df
#' @export
-print.epi_tibble = function(x, ...) {
- cat("An `epi_tibble` object, with metadata:\n")
+print.epi_df = function(x, ...) {
+ cat("An `epi_df` object, with metadata:\n")
cat(sprintf("* %-10s= %s\n", "geo_type", attributes(x)$metadata$geo_type))
cat(sprintf("* %-10s= %s\n", "time_type", attributes(x)$metadata$time_type))
cat(sprintf("* %-10s= %s\n", "issue", attributes(x)$metadata$issue))
@@ -239,28 +238,28 @@ print.epi_tibble = function(x, ...) {
NextMethod()
}
-#' @method head epi_tibble
+#' @method head epi_df
#' @importFrom utils head
#' @export
-head.epi_tibble = function(x, ...) {
+head.epi_df = function(x, ...) {
head(tibble::as_tibble(x), ...)
}
-#' Summarize `epi_tibble` object
+#' Summarize `epi_df` object
#'
-#' Prints a variety of summary statistics about the `epi_tibble` object, such as
+#' Prints a variety of summary statistics about the `epi_df` object, such as
#' the time range included and geographic coverage.
#'
-#' @param object The `epi_tibble` object.
+#' @param object The `epi_df` object.
#' @param ... Additional arguments, for compatibility with `summary()`.
#' Currently unused.
#' @return No return value; called only to print summary statistics.
#'
-#' @method summary epi_tibble
+#' @method summary epi_df
#' @importFrom stats median
#' @export
-summary.epi_tibble = function(object, ...) {
- cat("An `epi_tibble` object, with metadata:\n")
+summary.epi_df = function(object, ...) {
+ cat("An `epi_df` object, with metadata:\n")
cat(sprintf("* %-10s= %s\n", "geo_type", attributes(x)$metadata$geo_type))
cat(sprintf("* %-10s= %s\n", "time_type", attributes(x)$metadata$time_type))
cat(sprintf("* %-10s= %s\n", "issue", attributes(x)$metadata$issue))
@@ -273,29 +272,29 @@ summary.epi_tibble = function(object, ...) {
dplyr::summarize(median(.data$num)))))
}
-#' Group or ungroup `epi_tibble` object
+#' Group or ungroup `epi_df` object
#'
-#' Groups or ungroups an `epi_tibble`, preserving class and attributes.
+#' Groups or ungroups an `epi_df`, preserving class and attributes.
#'
-#' @method group_by epi_tibble
+#' @method group_by epi_df
#' @importFrom dplyr group_by
#' @export
-group_by.epi_tibble = function(x, ...) {
+group_by.epi_df = function(x, ...) {
metadata = attributes(x)$metadata
x = NextMethod()
- class(x) = c("epi_tibble", class(x))
+ class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
return(x)
}
-#' @method ungroup epi_tibble
-#' @rdname group_by.epi_tibble
+#' @method ungroup epi_df
+#' @rdname group_by.epi_df
#' @importFrom dplyr ungroup
#' @export
-ungroup.epi_tibble = function(x, ...) {
+ungroup.epi_df = function(x, ...) {
metadata = attributes(x)$metadata
x = NextMethod()
- class(x) = c("epi_tibble", class(x))
+ class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
return(x)
}
diff --git a/R/epitools.R b/R/epiprocess.R
similarity index 72%
rename from R/epitools.R
rename to R/epiprocess.R
index 731fbe04..0749647f 100644
--- a/R/epitools.R
+++ b/R/epiprocess.R
@@ -1,9 +1,9 @@
-#' epitools: Tools for basic signal processing in epidemiology
+#' epiprocess: Tools for basic signal processing in epidemiology
#'
#' This package introduces a common data structure for epidemiological data sets
#' measured over space and time, and offers associated utilities to perform
#' basic signal processing tasks.
#'
#' @docType package
-#' @name epitools
+#' @name epiprocess
NULL
diff --git a/R/outliers.R b/R/outliers.R
index c1a2572f..0bbcc58d 100644
--- a/R/outliers.R
+++ b/R/outliers.R
@@ -1,8 +1,10 @@
-#' Detect outliers in a variable in an `epi_tibble` object
+#' Detect outliers in a variable in an `epi_df` object
#'
-#' Applies one or more outlier detection methods to a variable in an
-#' `epi_tibble` object, and optionally aggregates the results to create
-#' consensus results.
+#' Applies one or more outlier detection methods to a variable in an `epi_df`
+#' object, and optionally aggregates the results to create consensus results.
+#' See the [outliers
+#' vignette](https://cmu-delphi.github.io/epitools/articles/outliers.html) for
+#' examples.
#'
#' @details Each outlier detection method, one per row of the passed `methods`
#' tibble, is a function that must take as its first two arguments `x` and
@@ -19,7 +21,7 @@
#' "stl", shorthand for `detect_outliers_stl()`, which detects outliers via an
#' STL decomposition.
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var The variable in `x` on which to run outlier detection.
#' @param methods A tibble specifying the method(s) to use for outlier
#' detection, with one row per method, and the following columns:
@@ -41,9 +43,9 @@
#' that setting `new_col_name` equal to an existing column name will overwrite
#' this column.
#'
-#' @return An `epi_tibble` object given by appending a new column to `x`, named
+#' @return An `epi_df` object given by appending a new column to `x`, named
#' according to the `new_col_name` argument, containing the outlier detection
-#' thresholds and replacement values from all detection methods.
+#' thresholds and replacement values from all detection methods.
#'
#' @importFrom dplyr group_modify mutate select
#' @importFrom purrr map pmap_dfc
@@ -57,8 +59,8 @@ detect_outliers = function(x, var,
abbr = "rm"),
combiner = c("median", "mean", "none"),
new_col_name = "outlier_info") {
- # Check we have an `epi_tibble` object
- if (!inherits(x, "epi_tibble")) abort("`x` must be of class `epi_tibble`.")
+ # Check we have an `epi_df` object
+ if (!inherits(x, "epi_df")) abort("`x` must be of class `epi_df`.")
# Check that we have a variable to do computations on
if (missing(var)) abort("`var` must be specified.")
@@ -79,7 +81,7 @@ detect_outliers = function(x, var,
new_col_name = new_col_name)
# Attach the class and metadata and return
-class(x) = c("epi_tibble", class(x))
+class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
return(x)
}
@@ -149,7 +151,7 @@ detect_outliers_one_grp = function(.data_group,
#' Detects outliers based on a distance from the rolling median specified in
#' terms of multiples of the rolling interquartile range (IQR).
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var The variable in `x` on which to run outlier detection.
#' @param n Number of time steps to use in the rolling window. Default is 21.
#' @param log_transform Should a log transform be applied before running outlier
@@ -240,7 +242,7 @@ detect_outliers_rolling_median = detect_outliers_rm
#' are exactly as in `detect_outliers_rm()`; refer to its help file for their
#' description.
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var The variable in `x` on which to run outlier detection.
#' @param n_trend Number of time steps to use in the rolling window for trend.
#' Default is 21.
diff --git a/R/pct_change.R b/R/pct_change.R
index 99d1c5fe..a2b47830 100644
--- a/R/pct_change.R
+++ b/R/pct_change.R
@@ -1,11 +1,11 @@
-#' Compute percentage change of a variable in an `epi_tibble` object
+#' Compute percentage change of a variable in an `epi_df` object
#'
-#' Computes the percentage change of a variable in an `epi_tibble` object. See
-#' the [percentage change
-#' vignette](https://cmu-delphi.github.io/epitools/articles/pct-change.html) for
-#' examples.
+#' Computes the percentage change of a variable in an `epi_df` object. See the
+#' [percentage change
+#' vignette](https://cmu-delphi.github.io/epitools/articles/pct_change.html) for
+#' examples.
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param var The variable in `x` whose percentage change values are to be
#' computed.
#' @param n Number of time steps to use in the running window. For example, if
@@ -33,7 +33,7 @@
#' metadata. Read the documentation for [epi_slide()] for more details.
#' @param na_rm Should missing values be removed before the computation? Default
#' is `TRUE`.
-#' @return An `epi_tibble` object given by appending a new column to `x`, named
+#' @return An `epi_df` object given by appending a new column to `x`, named
#' according to the `new_col_name` argument, containing the percentage change
#' values.
#'
diff --git a/R/slide.R b/R/slide.R
index 5cce4da0..0b7ef1da 100644
--- a/R/slide.R
+++ b/R/slide.R
@@ -1,8 +1,8 @@
-#' Slide a function over variables in an `epi_tibble` object
+#' Slide a function over variables in an `epi_df` object
#'
-#' Slides a given function over variables in an `epi_tibble` object. See the
-#' [slide vignette](https://cmu-delphi.github.io/epitools/articles/slide.html)
-#' for examples.
+#' Slides a given function over variables in an `epi_df` object. See the [slide
+#' vignette](https://cmu-delphi.github.io/epitools/articles/slide.html) for
+#' examples.
#'
#' @details To "slide" means to apply a function or formula over a running
#' window of `n` time steps, where the unit (the meaning of one time step) is
@@ -12,7 +12,7 @@
#' `time_step` argument (which if specified would override the default choice
#' based on the metadata).
#'
-#' @param x The `epi_tibble` object under consideration.
+#' @param x The `epi_df` object under consideration.
#' @param slide_fun Function or formula to slide over variables in `x`. To
#' "slide" means to apply a function or formula over a running window of `n`
#' time steps (where one time step is typically one day or one week; see
@@ -52,7 +52,7 @@
#' `time_type` is "day-time").
#' @param ... Additional arguments to pass to the function or formula specified
#' via `slide_fun`.
-#' @return An `epi_tibble` object given by appending a new column to `x`, named
+#' @return An `epi_df` object given by appending a new column to `x`, named
#' according to the `new_col_name` argument, containing the slide values.
#'
#' @details In order to slide a function or formula in a suitable grouped
@@ -67,8 +67,8 @@ epi_slide = function(x, slide_fun, n = 14, align = c("right", "center", "left"),
before, complete = FALSE, new_col_name = "slide_value",
new_col_type = c("dbl", "int", "lgl", "chr", "list"),
time_step, ...) {
- # Check we have an `epi_tibble` object
- if (!inherits(x, "epi_tibble")) abort("`x` must be of class `epi_tibble`.")
+ # Check we have an `epi_df` object
+ if (!inherits(x, "epi_df")) abort("`x` must be of class `epi_df`.")
# Which slide_index function?
new_col_type = match.arg(new_col_type)
@@ -125,7 +125,7 @@ epi_slide = function(x, slide_fun, n = 14, align = c("right", "center", "left"),
...)
# Attach the class and metadata and return
- class(x) = c("epi_tibble", class(x))
+ class(x) = c("epi_df", class(x))
attributes(x)$metadata = metadata
return(x)
}
diff --git a/_pkgdown.yml b/_pkgdown.yml
index 87fc22b7..a03fb348 100644
--- a/_pkgdown.yml
+++ b/_pkgdown.yml
@@ -13,9 +13,9 @@ articles:
desc: Basic usage and examples.
navbar: ~
contents:
- - epitools
+ - epiprocess
- slide
- - pct-change
+ - pct_change
- derivative
- correlation
- aggregation
diff --git a/docs/404.html b/docs/404.html
index 7e665f37..84502428 100644
--- a/docs/404.html
+++ b/docs/404.html
@@ -6,7 +6,7 @@
-Page not found (404) • epitools
+Page not found (404) • epiprocess
@@ -70,7 +70,7 @@
- epitools
+ epiprocess1.0.0
@@ -84,7 +84,7 @@
This package introduces a common data structure for epidemiological data sets measured over space and time, and offers associated utilities to perform basic signal processing tasks.
-
-
-Installing
-
This package is not on CRAN yet, so it can be installed using the devtools package:
Building the vignettes, such as this getting started guide, takes a significant amount of time. They are not included in the package by default. If you want to include vignettes, then use this modified command:
We’ll start by showing how to get data into epi_tibble format, which is just a tibble with a bit of special structure, and is the format assumed by all of the functions in the epitools package. An epi_tibble object has (at least) the following columns:
-
-
-geo_value: the geographic value associated with each measurement.
-
-time_value: the time value associated with each measurement.
-
-
Other columns can be considered as measured variables, which we also broadly refer to as signal variables. To learn more about the epi_tibble format, you can read the documentation for as.epi_tibble().
-
A data frame or tibble that has geo_value and time_value columns can be converted into an epi_tibble object, using the function as.epi_tibble(). As an example, we’ll look at daily cumulative COVID-19 cases for 4 states (CA, FL, NY, and TX) in the U.S., over a year spanning mid 2020 to mid 2021, using the covidcast package to fetch this data from the COVIDcast API.
As we can see, a data frame returned covidcast::covidcast_signal() has the columns required for an epi_tibble object (along with many others). A call to as.epi_tibble(), with further specification of some relevant metadata, brings the data frame into epi_tibble format.
## An `epi_tibble` object, with metadata:
-## * geo_type = state
-## * time_type = day
-## * issue = 2022-01-04
-##
-## Summary of space-time coverge:
-## * earliest time value = 2020-06-01
-## * latest time value = 2021-05-31
-## * median geo values per time value = 4
In general, an epi_tibble object must have following fields in its metadata:
-
-
-geo_type: the type for the geo values.
-
-time_type: the type for the time values.
-
-issue: the time value at which the given data set was issued.
-
-
In a typical case, the issue field represents the maximum of the issues of individual signal values measured in the data set; hence we would also say that the data set is comprised of all signal values observed “as of” the given issue in the metadata.
-
If any of the geo_type, time_type, and issue arguments are missing in a call to as.epi_tibble(), then this function will try to infer them from the passed object (The documentation for as.epi_tibble() gives more details.)
Data in epi_tibble format should be easy to work with downstream; in the other vignettes, we’ll walk through some basic signal processing tasks using functions provided in the epitools package; of course, we can also write custom code for for other downstream uses, like plotting, which is pretty easy to do ggplot2.
This package introduces a common data structure for epidemiological data sets measured over space and time, and offers associated utilities to perform basic signal processing tasks.
Computes lagged correlations between variables in an epi_tibble object,
-allowing for grouping by geo value, time value, or any other variables. See
-the correlations vignette
-for examples.
Time shifts to consider for the two variables, respectively,
-before computing correlations. Negative shifts translate into in a lag
-value and positive shifts into a lead value; for example, if dt = -1,
-then the new value on June 2 is the original value on June 1; if dt = 1,
-then the new value on June 2 is the original value on June 3; if dt = 0,
-then the values are left as is. Default is 0 for both dt1 and dt2. Note
-that the time shifts are always performed per geo value; see details.
-
-
-
by
-
The variable(s) to group by, for the correlation computation. If
-geo_value, the default, then correlations are computed for each geo
-value, over all time; if time_value, then correlations are computed for
-each time, over all geo values. A grouping can also be any specified using
-number of columns of x; for example, we can use by = c(geo_value, age_group), assuming x has a column age_group, in order to compute
-correlations for each pair of geo value and age group. To omit a grouping
-entirely, use by = NULL. Note that the grouping here is always applied
-after the time shifts; see details.
-
-
-
use, method
-
Arguments to pass to cor(), with "na.or.complete" the
-default for use (different than cor()) and "pearson" the default for
-method (same as cor()).
-
-
-
-
Value
-
-
An tibble with the grouping columns first (geo_value, time_value,
-or possibly others), and then a column cor, which gives the correlation.
-
Details
-
-
Time shifts are always performed first, grouped by geo value (this
-way they amount to shifting each individual time series). After this, the
-geo grouping is removed, and the grouping specified in the by argument is
-applied. Then, correlations are computed.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
diff --git a/docs/reference/detect_outliers.html b/docs/reference/detect_outliers.html
index e2680f5c..4929f5fd 100644
--- a/docs/reference/detect_outliers.html
+++ b/docs/reference/detect_outliers.html
@@ -6,7 +6,7 @@
-Detect outliers in a variable in an <code>epi_tibble</code> object — detect_outliers • epitools
+Detect outliers in a variable in an <code>epi_df</code> object — detect_outliers • epiprocess
@@ -39,10 +39,11 @@
-
-
+
+
@@ -73,7 +74,7 @@
- epitools
+ epiprocess1.0.0
Applies one or more outlier detection methods to a variable in an
-epi_tibble object, and optionally aggregates the results to create
-consensus results.
+
Applies one or more outlier detection methods to a variable in an epi_df
+object, and optionally aggregates the results to create consensus results.
+See the outliers vignette for
+examples.
detect_outliers(
@@ -164,7 +169,7 @@
Arg
x
-
The epi_tibble object under consideration.
+
The epi_df object under consideration.
var
@@ -203,7 +208,7 @@
Arg
Value
-
An epi_tibble object given by appending a new column to x, named
+
An epi_df object given by appending a new column to x, named
according to the new_col_name argument, containing the outlier detection
thresholds and replacement values from all detection methods.
Details
diff --git a/docs/reference/detect_outliers_rm.html b/docs/reference/detect_outliers_rm.html
index 225a9618..3eb8bedb 100644
--- a/docs/reference/detect_outliers_rm.html
+++ b/docs/reference/detect_outliers_rm.html
@@ -6,7 +6,7 @@
-Detect outliers based on a rolling median — detect_outliers_rm • epitools
+Detect outliers based on a rolling median — detect_outliers_rm • epiprocess
@@ -72,7 +72,7 @@
- epitools
+ epiprocess1.0.0
@@ -86,7 +86,7 @@
This package introduces a common data structure for epidemiological data sets
-measured over space and time, and offers associated utilities to perform
-basic signal processing tasks.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
diff --git a/docs/reference/estimate_deriv.html b/docs/reference/estimate_deriv.html
index 567eefc1..555b449f 100644
--- a/docs/reference/estimate_deriv.html
+++ b/docs/reference/estimate_deriv.html
@@ -6,7 +6,7 @@
-Estimate derivatives of a variable in an <code>epi_tibble</code> object — estimate_deriv • epitools
+Estimate derivatives of a variable in an <code>epi_df</code> object — estimate_deriv • epiprocess
@@ -39,10 +39,10 @@
-
-
+
+
@@ -73,7 +73,7 @@
- epitools
+ epiprocess1.0.0
@@ -87,7 +87,7 @@
Estimates derivatives of a variable in an epi_tibble object, using a local
-(in time) linear regression or alternative smoother. See the estimating derivatives vignette
-for examples.
+
Estimates derivatives of a variable in an epi_df object, using a local (in
+time) linear regression or alternative smoothing method. See the estimating derivatives vignette for
+examples.
estimate_deriv(
@@ -172,7 +175,7 @@
Arg
x
-
The epi_tibble object under consideration.
+
The epi_df object under consideration.
var
@@ -256,7 +259,7 @@
Arg
Value
-
An epi_tibble object given by appending a new column to x, named
+
An epi_df object given by appending a new column to x, named
according to the new_col_name argument, containing the derivative values.
diff --git a/index.md b/index.md
index ae7e95cd..1e4914e0 100644
--- a/index.md
+++ b/index.md
@@ -1,4 +1,4 @@
-# epitools
+# epiprocess
This package introduces a common data structure for epidemiological data sets
measured over space and time, and offers associated utilities to perform basic
diff --git a/man/as.epi_tibble.Rd b/man/as.epi_tibble.Rd
deleted file mode 100644
index ff82b568..00000000
--- a/man/as.epi_tibble.Rd
+++ /dev/null
@@ -1,142 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/epi_tibble.R
-\name{as.epi_tibble}
-\alias{as.epi_tibble}
-\alias{as.epi_tibble.epi_tibble}
-\alias{as.epi_tibble.tibble}
-\alias{as.epi_tibble.data.frame}
-\title{Convert data to \code{epi_tibble} format}
-\usage{
-as.epi_tibble(x, ...)
-
-\method{as.epi_tibble}{epi_tibble}(x, ...)
-
-\method{as.epi_tibble}{tibble}(x, geo_type, time_type, issue, additional_metadata = list(), ...)
-
-\method{as.epi_tibble}{data.frame}(x, geo_type, time_type, issue, additional_metadata = list(), ...)
-}
-\arguments{
-\item{x}{The object to be converted. See the methods section below for
-details on formatting of each input type.}
-
-\item{...}{Additional arguments passed to methods.}
-
-\item{geo_type}{The type for the geo values. If missing, then the function
-will attempt to infer it from the geo values present; if this fails, then
-it will be set to "custom".}
-
-\item{time_type}{The type for the time values. If missing, then the function
-will attempt to infer it from the time values present; if this fails, then
-it will be set to "custom".}
-
-\item{issue}{Issue to use for this data. If missing, then the function will
-attempt to infer it from the passed object \code{x}; if this fails, then the
-current day-time will be used.}
-
-\item{additional_metadata}{List of additional metadata to attach to the
-\code{epi_tibble} object. All objects will have \code{time_type}, \code{geo_type}, and
-\code{issue} fields; named entries from the passed list or will be included as
-well.}
-}
-\value{
-An \code{epi_tibble} object.
-}
-\description{
-Converts a data frame or tibble into a format that is consistent with the
-\code{epi_tibble} class, ensuring that it has a certain minimal set of columns,
-and that it has certain minimal metadata.
-}
-\details{
-An \code{epi_tibble} is a tibble with (at least) the following columns:
-\itemize{
-\item \code{geo_value}: the geographic value associated with each measurement.
-\item \code{time_value}: the time value associated with each measurement.
-}
-
-Other columns can be considered as measured variables, which we also broadly
-refer to as signal variables. An \code{epi_tibble} object also has metadata with
-(at least) the following fields:
-\itemize{
-\item \code{geo_type}: the type for the geo values.
-\item \code{time_type}: the type for the time values.
-\item \code{issue}: the time value at which the given data set was issued.
-}
-
-The first two fields above, \code{geo_type} and \code{time_type}, can usually be
-inferred from the \code{geo_value} and \code{time_value} columns, respectively. The
-last field above, \code{issue}, is the most unique to the \code{epi_tibble} format.
-In a typical case, this represents the maximum of the issues of individual
-signal values measured in the data set; hence we would also say that the
-data set is comprised of all signal values observed "as of" the given issue
-in the metadata.
-
-Metadata for an \code{epi_tibble} object \code{x} can be accessed (and altered) via
-\code{attributes(x)$metadata}. More information on geo types, time types, and
-issues is given below.
-}
-\section{Methods (by class)}{
-\itemize{
-\item \code{epi_tibble}: Simply returns the \code{epi_tibble} object unchanged.
-
-\item \code{tibble}: The input tibble \code{x} must contain the columns
-\code{geo_value} and \code{time_value}. All other columns will be preserved as is,
-and treated as measured variables. If \code{issue} is missing, then the function
-will look for \code{issue} as a column of \code{x}, or as a field in its metadata
-(stored in its attributes), to infer the issue; if this fails, then the
-current day-time will be used.
-
-\item \code{data.frame}: The input data frame \code{x} must contain the columns
-\code{geo_value} and \code{time_value}. All other columns will be preserved as is,
-and treated as measured variables. If \code{issue} is missing, then the function
-will look for \code{issue} as a column of \code{x}, or as a field in its metadata
-(stored in its attributes), to infer the issue; if this fails, then the
-current day-time will be used.
-}}
-
-\section{Geo types}{
-
-The following geo types are supported in an \code{epi_tibble}. Their geo coding
-(specification of geo values for each geo type) is also described below.
-\itemize{
-\item \code{"county"}: each observation corresponds to a U.S. county; coded by 5-digit
-FIPS code.
-\item \code{"hrr"}: each observation corresponds to a U.S. hospital referral region
-(designed to represent regional healthcare markets); there are 306 HRRs in
-the U.S; coded by number (nonconsecutive, between 1 and 457).
-\item \code{"state"}: each observation corresponds to a U.S. state; coded by 2-digit
-postal abbreviation (lowercase);
-note that Puerto Rico is "pr" and Washington D.C. is "dc".
-\item \code{"hhs"}: each observation corresponds to a U.S. HHS region; coded by number
-(consecutive, between 1 and 10).
-\item \code{"nation"}: each observation corresponds to a country; coded by ISO 31661-
-alpha-2 country codes (lowercase).
-}
-
-The above geo types come with aggregation utilities in the package, \emph{todo:
-refer to relevant functionality, vignette, and so on}. An unrecognizable
-geo type is labeled as "custom".
-}
-
-\section{Time types}{
-
-The following time types are supported in an \code{epi_tibble}. Their time coding
-(specification of time values for each time type) is also described below.
-\itemize{
-\item \code{"day-time"}: each observation corresponds to a time on a given day (measured
-to the second); coded as a \code{POSIXct} object, as in \code{as.POSIXct("2020-06-09 18:45:40")}.
-\item \code{"day"}: each observation corresponds to a day; coded as a \code{Date} object,
-as in \code{as.Date("2020-06-09")}.
-\item \code{"week"}: each observation corresponds to a week; the alignment can be
-arbitrary (as to whether a week starts on a Monday, Tuesday, etc.; the
-U.S. CDC definition of an epidemiological week starts on a Sunday); coded
-as a \code{Date} object, representing the start date of week.
-}
-
-An unrecognisable time type is labeled as "custom".
-}
-
-\section{Issues}{
-
-todo
-}
-
diff --git a/man/cor_lagged.Rd b/man/cor_lagged.Rd
deleted file mode 100644
index 9fc4a163..00000000
--- a/man/cor_lagged.Rd
+++ /dev/null
@@ -1,59 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/correlation.R
-\name{cor_lagged}
-\alias{cor_lagged}
-\title{Compute lagged correlations between variables in an \code{epi_tibble} object}
-\usage{
-cor_lagged(
- x,
- var1,
- var2,
- dt1 = 0,
- dt2 = 0,
- by = geo_value,
- use = "na.or.complete",
- method = c("pearson", "kendall", "spearman")
-)
-}
-\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
-
-\item{var1, var2}{The variables in \code{x} to correlate.}
-
-\item{dt1, dt2}{Time shifts to consider for the two variables, respectively,
-before computing correlations. Negative shifts translate into in a lag
-value and positive shifts into a lead value; for example, if \code{dt = -1},
-then the new value on June 2 is the original value on June 1; if \code{dt = 1},
-then the new value on June 2 is the original value on June 3; if \code{dt = 0},
-then the values are left as is. Default is 0 for both \code{dt1} and \code{dt2}. Note
-that the time shifts are always performed \emph{per geo value}; see details.}
-
-\item{by}{The variable(s) to group by, for the correlation computation. If
-\code{geo_value}, the default, then correlations are computed for each geo
-value, over all time; if \code{time_value}, then correlations are computed for
-each time, over all geo values. A grouping can also be any specified using
-number of columns of \code{x}; for example, we can use \code{by = c(geo_value, age_group)}, assuming \code{x} has a column \code{age_group}, in order to compute
-correlations for each pair of geo value and age group. To omit a grouping
-entirely, use \code{by = NULL}. Note that the grouping here is always applied
-\emph{after} the time shifts; see details.}
-
-\item{use, method}{Arguments to pass to \code{cor()}, with "na.or.complete" the
-default for \code{use} (different than \code{cor()}) and "pearson" the default for
-\code{method} (same as \code{cor()}).}
-}
-\value{
-An tibble with the grouping columns first (\code{geo_value}, \code{time_value},
-or possibly others), and then a column \code{cor}, which gives the correlation.
-}
-\description{
-Computes lagged correlations between variables in an \code{epi_tibble} object,
-allowing for grouping by geo value, time value, or any other variables. See
-the \href{https://cmu-delphi.github.io/epitools/articles/correlations.html}{correlations vignette}
-for examples.
-}
-\details{
-Time shifts are always performed first, grouped by geo value (this
-way they amount to shifting each individual time series). After this, the
-geo grouping is removed, and the grouping specified in the \code{by} argument is
-applied. Then, correlations are computed.
-}
diff --git a/man/detect_outliers.Rd b/man/detect_outliers.Rd
index 07101c7c..3c624fe3 100644
--- a/man/detect_outliers.Rd
+++ b/man/detect_outliers.Rd
@@ -2,7 +2,7 @@
% Please edit documentation in R/outliers.R
\name{detect_outliers}
\alias{detect_outliers}
-\title{Detect outliers in a variable in an \code{epi_tibble} object}
+\title{Detect outliers in a variable in an \code{epi_df} object}
\usage{
detect_outliers(
x,
@@ -13,7 +13,7 @@ detect_outliers(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{var}{The variable in \code{x} on which to run outlier detection.}
@@ -42,14 +42,15 @@ that setting \code{new_col_name} equal to an existing column name will overwrite
this column.}
}
\value{
-An \code{epi_tibble} object given by appending a new column to \code{x}, named
+An \code{epi_df} object given by appending a new column to \code{x}, named
according to the \code{new_col_name} argument, containing the outlier detection
thresholds and replacement values from all detection methods.
}
\description{
-Applies one or more outlier detection methods to a variable in an
-\code{epi_tibble} object, and optionally aggregates the results to create
-consensus results.
+Applies one or more outlier detection methods to a variable in an \code{epi_df}
+object, and optionally aggregates the results to create consensus results.
+See the \href{https://cmu-delphi.github.io/epitools/articles/outliers.html}{outliers vignette} for
+examples.
}
\details{
Each outlier detection method, one per row of the passed \code{methods}
diff --git a/man/detect_outliers_rm.Rd b/man/detect_outliers_rm.Rd
index df8c3b3d..a10be46d 100644
--- a/man/detect_outliers_rm.Rd
+++ b/man/detect_outliers_rm.Rd
@@ -16,7 +16,7 @@ detect_outliers_rm(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{var}{The variable in \code{x} on which to run outlier detection.}
diff --git a/man/detect_outliers_stl.Rd b/man/detect_outliers_stl.Rd
index fc1eb718..c2befecb 100644
--- a/man/detect_outliers_stl.Rd
+++ b/man/detect_outliers_stl.Rd
@@ -19,7 +19,7 @@ detect_outliers_stl(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{var}{The variable in \code{x} on which to run outlier detection.}
diff --git a/man/epi_slide.Rd b/man/epi_slide.Rd
index 1591cea5..93621a40 100644
--- a/man/epi_slide.Rd
+++ b/man/epi_slide.Rd
@@ -2,7 +2,7 @@
% Please edit documentation in R/slide.R
\name{epi_slide}
\alias{epi_slide}
-\title{Slide a function over variables in an \code{epi_tibble} object}
+\title{Slide a function over variables in an \code{epi_df} object}
\usage{
epi_slide(
x,
@@ -18,7 +18,7 @@ epi_slide(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{slide_fun}{Function or formula to slide over variables in \code{x}. To
"slide" means to apply a function or formula over a running window of \code{n}
@@ -67,13 +67,12 @@ only be meaningful if \code{time_value} is of class \code{POSIXct}, that is, if
via \code{slide_fun}.}
}
\value{
-An \code{epi_tibble} object given by appending a new column to \code{x}, named
+An \code{epi_df} object given by appending a new column to \code{x}, named
according to the \code{new_col_name} argument, containing the slide values.
}
\description{
-Slides a given function over variables in an \code{epi_tibble} object. See the
-\href{https://cmu-delphi.github.io/epitools/articles/slide.html}{slide vignette}
-for examples.
+Slides a given function over variables in an \code{epi_df} object. See the \href{https://cmu-delphi.github.io/epitools/articles/slide.html}{slide vignette} for
+examples.
}
\details{
To "slide" means to apply a function or formula over a running
diff --git a/man/epitools.Rd b/man/epitools.Rd
deleted file mode 100644
index 44cf6fab..00000000
--- a/man/epitools.Rd
+++ /dev/null
@@ -1,11 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/epitools.R
-\docType{package}
-\name{epitools}
-\alias{epitools}
-\title{epitools: Tools for basic signal processing in epidemiology}
-\description{
-This package introduces a common data structure for epidemiological data sets
-measured over space and time, and offers associated utilities to perform
-basic signal processing tasks.
-}
diff --git a/man/estimate_deriv.Rd b/man/estimate_deriv.Rd
index dacf3647..ef2a5eb1 100644
--- a/man/estimate_deriv.Rd
+++ b/man/estimate_deriv.Rd
@@ -2,7 +2,7 @@
% Please edit documentation in R/derivative.R
\name{estimate_deriv}
\alias{estimate_deriv}
-\title{Estimate derivatives of a variable in an \code{epi_tibble} object}
+\title{Estimate derivatives of a variable in an \code{epi_df} object}
\usage{
estimate_deriv(
x,
@@ -21,7 +21,7 @@ estimate_deriv(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{var}{The variable in \code{x} whose derivatives are to be estimated.}
@@ -78,13 +78,13 @@ is \code{TRUE}.}
derivatives. See details below.}
}
\value{
-An \code{epi_tibble} object given by appending a new column to \code{x}, named
+An \code{epi_df} object given by appending a new column to \code{x}, named
according to the \code{new_col_name} argument, containing the derivative values.
}
\description{
-Estimates derivatives of a variable in an \code{epi_tibble} object, using a local
-(in time) linear regression or alternative smoother. See the \href{https://cmu-delphi.github.io/epitools/articles/derivatives.html}{estimating derivatives vignette}
-for examples.
+Estimates derivatives of a variable in an \code{epi_df} object, using a local (in
+time) linear regression or alternative smoothing method. See the \href{https://cmu-delphi.github.io/epitools/articles/derivative.html}{estimating derivatives vignette} for
+examples.
}
\details{
Derivatives are estimated using:
diff --git a/man/group_by.epi_tibble.Rd b/man/group_by.epi_tibble.Rd
deleted file mode 100644
index be50410d..00000000
--- a/man/group_by.epi_tibble.Rd
+++ /dev/null
@@ -1,14 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/epi_tibble.R
-\name{group_by.epi_tibble}
-\alias{group_by.epi_tibble}
-\alias{ungroup.epi_tibble}
-\title{Group or ungroup \code{epi_tibble} object}
-\usage{
-\method{group_by}{epi_tibble}(x, ...)
-
-\method{ungroup}{epi_tibble}(x, ...)
-}
-\description{
-Groups or ungroups an \code{epi_tibble}, preserving class and attributes.
-}
diff --git a/man/pct_change.Rd b/man/pct_change.Rd
index 9de92ac3..27863a7f 100644
--- a/man/pct_change.Rd
+++ b/man/pct_change.Rd
@@ -2,7 +2,7 @@
% Please edit documentation in R/pct_change.R
\name{pct_change}
\alias{pct_change}
-\title{Compute percentage change of a variable in an \code{epi_tibble} object}
+\title{Compute percentage change of a variable in an \code{epi_df} object}
\usage{
pct_change(
x,
@@ -17,7 +17,7 @@ pct_change(
)
}
\arguments{
-\item{x}{The \code{epi_tibble} object under consideration.}
+\item{x}{The \code{epi_df} object under consideration.}
\item{var}{The variable in \code{x} whose percentage change values are to be
computed.}
@@ -55,13 +55,13 @@ metadata. Read the documentation for \code{\link[=epi_slide]{epi_slide()}} for m
is \code{TRUE}.}
}
\value{
-An \code{epi_tibble} object given by appending a new column to \code{x}, named
+An \code{epi_df} object given by appending a new column to \code{x}, named
according to the \code{new_col_name} argument, containing the percentage change
values.
}
\description{
-Computes the percentage change of a variable in an \code{epi_tibble} object. See
-the \href{https://cmu-delphi.github.io/epitools/articles/pct-change.html}{percentage change vignette} for
+Computes the percentage change of a variable in an \code{epi_df} object. See the
+\href{https://cmu-delphi.github.io/epitools/articles/pct_change.html}{percentage change vignette} for
examples.
}
\seealso{
diff --git a/man/print.epi_tibble.Rd b/man/print.epi_tibble.Rd
deleted file mode 100644
index 297300bb..00000000
--- a/man/print.epi_tibble.Rd
+++ /dev/null
@@ -1,21 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/epi_tibble.R
-\name{print.epi_tibble}
-\alias{print.epi_tibble}
-\title{Print \code{epi_tibble} object}
-\usage{
-\method{print}{epi_tibble}(x, ...)
-}
-\arguments{
-\item{x}{The \code{epi_tibble} object.}
-
-\item{...}{Additional arguments passed to \code{print.tibble()} to print the
-data.}
-}
-\value{
-The \code{epi_tibble} object, unchanged.
-}
-\description{
-Prints a brief summary of the \code{epi_tibble} object, then prints the underlying
-tibble.
-}
diff --git a/man/summary.epi_tibble.Rd b/man/summary.epi_tibble.Rd
deleted file mode 100644
index d82a8523..00000000
--- a/man/summary.epi_tibble.Rd
+++ /dev/null
@@ -1,21 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/epi_tibble.R
-\name{summary.epi_tibble}
-\alias{summary.epi_tibble}
-\title{Summarize \code{epi_tibble} object}
-\usage{
-\method{summary}{epi_tibble}(object, ...)
-}
-\arguments{
-\item{object}{The \code{epi_tibble} object.}
-
-\item{...}{Additional arguments, for compatibility with \code{summary()}.
-Currently unused.}
-}
-\value{
-No return value; called only to print summary statistics.
-}
-\description{
-Prints a variety of summary statistics about the \code{epi_tibble} object, such as
-the time range included and geographic coverage.
-}
diff --git a/vignettes/correlation.Rmd b/vignettes/correlation.Rmd
index cf6e743e..25cbac53 100644
--- a/vignettes/correlation.Rmd
+++ b/vignettes/correlation.Rmd
@@ -7,8 +7,8 @@ vignette: >
%\VignetteEncoding{UTF-8}
---
-The `epitools` package provides some simple functionality for computing lagged
-correlations between two signals, over space or time, via `cor_lagged()`. This
+The `epiprocess` package provides some simple functionality for computing lagged
+correlations between two signals, over space or time, via `epi_cor()`. This
function is really just a convenience wrapper over some basic commands: it first
performs any specified time shifts, grouped by geo value (this way, it shifts
each of the individual time series); then, it removes the geo grouping and
@@ -19,7 +19,7 @@ and death rates, smoothed using a 7-day trailing average.
```{r, message = FALSE}
library(covidcast)
-library(epitools)
+library(epiprocess)
library(dplyr)
start_day <- "2020-03-01"
@@ -43,7 +43,7 @@ y <- covidcast_signal(data_source = "jhu-csse",
x <- x %>%
full_join(y, by = c("geo_value", "time_value")) %>%
- as.epi_tibble() %>%
+ as.epi_df() %>%
group_by(geo_value) %>%
epi_slide(slide_fun = ~ Mean(.x$case_rates), n = 7,
new_col_name = "case_rates") %>%
@@ -54,21 +54,21 @@ x <- x %>%
## Correlations grouped by time
-The `cor_lagged()` function operates on an `epi_tibble` object, and it requires
-further specification of the variables to correlate, in its next two arguments
-(`var1` and `var2`).
+The `epi_cor()` function operates on an `epi_df` object, and it requires further
+specification of the variables to correlate, in its next two arguments (`var1`
+and `var2`).
In general, we can specify any grouping variable (or combination of variables)
-in a call to `cor_lagged()`, via the `by` argument. This potentially leads to
-many ways to compute correlations. There are always at least two ways to compute
-correlations in an `epi_tibble`: grouping by time value, and by geo value. The
+in a call to `epi_cor()`, via the `by` argument. This potentially leads to many
+ways to compute correlations. There are always at least two ways to compute
+correlations in an `epi_df`: grouping by time value, and by geo value. The
former is obtained by setting `by = time_value`.
```{r, message = FALSE, warning = FALSE}
library(ggplot2)
theme_set(theme_bw())
-z1 <- cor_lagged(x, case_rates, death_rates, by = time_value)
+z1 <- epi_cor(x, case_rates, death_rates, by = time_value)
ggplot(z1, aes(x = time_value, y = cor)) +
geom_line() +
@@ -81,21 +81,21 @@ rates linearly associated, across the U.S. states?". We might be interested in
broadening this question, instead asking: "on any given day, do higher case
rates tend to associate with higher death rates?", removing the dependence on a
linear relationship. The latter can be addressed using Spearman correlation,
-accomplished by setting `method = "spearman"` in the call to `cor_lagged()`.
+accomplished by setting `method = "spearman"` in the call to `epi_cor()`.
Spearman correlation is highly robust and invariant to monotone transformations.
## Lagged correlations
We might also be interested in how case rates associate with death rates in the
-*future*. Using the `dt1` parameter in `cor_lagged()`, we can lag case rates
-back any number of days we want, before calculating correlations. Below, we set
-`dt1 = -10`. This means that `var1 = case_rates` will be lagged by 10 days, so
-that case rates on June 1st will be correlated with death rates on June 11th.
-(It might also help to think of it this way: death rates on a certain day will
-be correlated with case rates at an offset of -10 days.)
+*future*. Using the `dt1` parameter in `epi_cor()`, we can lag case rates back
+any number of days we want, before calculating correlations. Below, we set `dt1
+= -10`. This means that `var1 = case_rates` will be lagged by 10 days, so that
+case rates on June 1st will be correlated with death rates on June 11th. (It
+might also help to think of it this way: death rates on a certain day will be
+correlated with case rates at an offset of -10 days.)
```{r, message = FALSE, warning = FALSE}
-z2 <- cor_lagged(x, case_rates, death_rates, by = time_value, dt1 = -10)
+z2 <- epi_cor(x, case_rates, death_rates, by = time_value, dt1 = -10)
z <- rbind(z1 %>% mutate(lag = 0),
z2 %>% mutate(lag = 10)) %>%
@@ -118,8 +118,8 @@ geo_value`. We'll again look at correlations both for 0- and 10-day lagged case
rates.
```{r, message = FALSE, warning = FALSE}
-z1 <- cor_lagged(x, case_rates, death_rates, by = geo_value)
-z2 <- cor_lagged(x, case_rates, death_rates, by = geo_value, dt1 = -10)
+z1 <- epi_cor(x, case_rates, death_rates, by = geo_value)
+z2 <- epi_cor(x, case_rates, death_rates, by = geo_value, dt1 = -10)
z <- rbind(z1 %>% mutate(lag = 0),
z2 %>% mutate(lag = 10)) %>%
@@ -143,7 +143,7 @@ library(purrr)
lags = 0:35
z <- map_dfr(lags, function(lag) {
- cor_lagged(x, case_rates, death_rates, by = geo_value, dt1 = -lag) %>%
+ epi_cor(x, case_rates, death_rates, by = geo_value, dt1 = -lag) %>%
mutate(lag = lag)
})
diff --git a/vignettes/derivative.Rmd b/vignettes/derivative.Rmd
index 3bb6d12f..c2ba9d05 100644
--- a/vignettes/derivative.Rmd
+++ b/vignettes/derivative.Rmd
@@ -14,7 +14,7 @@ per 100,000 people), smoothed using a 7-day trailing average.
```{r, message = FALSE}
library(covidcast)
-library(epitools)
+library(epiprocess)
library(dplyr)
x <- covidcast_signal(data_source = "jhu-csse",
@@ -23,7 +23,7 @@ x <- covidcast_signal(data_source = "jhu-csse",
end_day = "2021-05-31",
geo_type = "state",
geo_values = "fl") %>%
- as.epi_tibble() %>%
+ as.epi_df() %>%
rename(case_rates = value) %>%
select(geo_value, time_value, case_rates) %>%
epi_slide(slide_fun = ~ Mean(.x$case_rates), n = 7,
@@ -32,13 +32,13 @@ x <- covidcast_signal(data_source = "jhu-csse",
## Estimating derivatives via linear regression
-The `estimate_deriv()` function operates on a `epi_tibble` object, and aside
-from this, takes three main arguments: `var`, the variable whose derivatives are
-to be estimated; `method`, the name of the method to use for derivative
-estimation; and `n`, the length of the local window (here, in days) to use in
-training the given method. Here we use `method = "lin"`, the default, which
-estimates the derivative using the slope from a simple linear regression, and
-`n = 14`, also the default.
+The `estimate_deriv()` function operates on a `epi_df` object, and aside from
+this, takes three main arguments: `var`, the variable whose derivatives are to
+be estimated; `method`, the name of the method to use for derivative estimation;
+and `n`, the length of the local window (here, in days) to use in training the
+given method. Here we use `method = "lin"`, the default, which estimates the
+derivative using the slope from a simple linear regression, and `n = 14`, also
+the default.
```{r}
x <- estimate_deriv(x, case_rates, method = "lin", n = 14)
diff --git a/vignettes/epitools.Rmd b/vignettes/epiprocess.Rmd
similarity index 67%
rename from vignettes/epitools.Rmd
rename to vignettes/epiprocess.Rmd
index c1503781..2e5952af 100644
--- a/vignettes/epitools.Rmd
+++ b/vignettes/epiprocess.Rmd
@@ -1,8 +1,8 @@
---
-title: Get started with `epitools`
+title: Get started with `epiprocess`
output: rmarkdown::html_vignette
vignette: >
- %\VignetteIndexEntry{Get started with epitools}
+ %\VignetteIndexEntry{Get started with epiprocess}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
@@ -17,7 +17,7 @@ This package is not on CRAN yet, so it can be installed using the
[`devtools`](https://cran.r-project.org/package=devtools) package:
```{r, eval = FALSE}
-devtools::install_github("cmu-delphi/epitools", ref = "main")
+devtools::install_github("cmu-delphi/epiprocess", ref = "main")
```
Building the vignettes, such as this getting started guide, takes a significant
@@ -25,31 +25,31 @@ amount of time. They are not included in the package by default. If you want to
include vignettes, then use this modified command:
```{r, eval = FALSE}
-devtools::install_github("cmu-delphi/epitools", ref = "main",
+devtools::install_github("cmu-delphi/epiprocess", ref = "main",
build_vignettes = TRUE, dependencies = TRUE)
```
-## Getting data into `epi_tibble` format
+## Getting data into `epi_df` format
-We'll start by showing how to get data into `epi_tibble` format, which is just a
+We'll start by showing how to get data into `epi_df` format, which is just a
tibble with a bit of special structure, and is the format assumed by all of the
-functions in the `epitools` package. An `epi_tibble` object has (at least) the
-following columns:
+functions in the `epiprocess` package. An `epi_df` object has (at least) the
+following columns:
* `geo_value`: the geographic value associated with each measurement.
* `time_value`: the time value associated with each measurement.
Other columns can be considered as measured variables, which we also broadly
-refer to as signal variables. To learn more about the `epi_tibble` format, you
-can read the documentation for `as.epi_tibble()`.
+refer to as signal variables. To learn more about the `epi_df` format, you can
+read the documentation for `as.epi_df()`.
-A data frame or tibble that has `geo_value` and `time_value` columns can be
-converted into an `epi_tibble` object, using the function `as.epi_tibble()`. As
-an example, we'll look at daily cumulative COVID-19 cases for 4 states (CA, FL,
-NY, and TX) in the U.S., over a year spanning mid 2020 to mid 2021, using the
-[`covidcast`](https://cmu-delphi.github.io/covidcast/covidcastR/)
-package to fetch this data from the [COVIDcast
-API](https://cmu-delphi.github.io/delphi-epidata/api/covidcast.html/).
+A data frame or tibble that has `geo_value` and `time_value` columns can be
+converted into an `epi_df` object, using the function `as.epi_df()`. As an
+example, we'll look at daily cumulative COVID-19 cases for 4 states (CA, FL, NY,
+and TX) in the U.S., over a year spanning mid 2020 to mid 2021, using the
+[`covidcast`](https://cmu-delphi.github.io/covidcast/covidcastR/) package to
+fetch this data from the [COVIDcast
+API](https://cmu-delphi.github.io/delphi-epidata/api/covidcast.html/).
```{r, message = FALSE}
library(covidcast)
@@ -65,19 +65,19 @@ class(case_data)
colnames(case_data)
```
-As we can see, a data frame returned `covidcast::covidcast_signal()` has the
-columns required for an `epi_tibble` object (along with many others). A call to
-`as.epi_tibble()`, with further specification of some relevant metadata, brings
-the data frame into `epi_tibble` format.
+As we can see, a data frame returned `covidcast::covidcast_signal()` has the
+columns required for an `epi_df` object (along with many others). A call to
+`as.epi_df()`, with further specification of some relevant metadata, brings the
+data frame into `epi_df` format.
```{r, message = FALSE}
-library(epitools)
+library(epiprocess)
library(dplyr)
-x <- as.epi_tibble(case_data,
- geo_type = "state",
- time_type = "day",
- issue = max(case_data$issue)) %>%
+x <- as.epi_df(case_data,
+ geo_type = "state",
+ time_type = "day",
+ issue = max(case_data$issue)) %>%
rename("cases_cum" = value) %>%
select(geo_value, time_value, cases_cum)
@@ -89,7 +89,7 @@ attributes(x)$metadata
## Some info about required metadata
-In general, an `epi_tibble` object must have following fields in its metadata:
+In general, an `epi_df` object must have following fields in its metadata:
* `geo_type`: the type for the geo values.
* `time_type`: the type for the time values.
@@ -100,24 +100,24 @@ individual signal values measured in the data set; hence we would also say that
the data set is comprised of all signal values observed "as of" the given issue
in the metadata.
-If any of the `geo_type`, `time_type`, and `issue` arguments are missing in a
-call to `as.epi_tibble()`, then this function will try to infer them from the
-passed object (The documentation for `as.epi_tibble()` gives more details.)
+If any of the `geo_type`, `time_type`, and `issue` arguments are missing in a
+call to `as.epi_df()`, then this function will try to infer them from the passed
+object (The documentation for `as.epi_df()` gives more details.)
```{r}
-x <- as.epi_tibble(case_data) %>%
+x <- as.epi_df(case_data) %>%
rename("cases_cum" = value) %>%
select(geo_value, time_value, cases_cum)
attributes(x)$metadata
```
-## Working with `epi_tibble` objects downstream
+## Working with `epi_df` objects downstream
-Data in `epi_tibble` format should be easy to work with downstream; in the other
-vignettes, we'll walk through some basic signal processing tasks using functions
-provided in the `epitools` package; of course, we can also write custom code for
-for other downstream uses, like plotting, which is pretty easy to do `ggplot2`.
+Data in `epi_df` format should be easy to work with downstream; in the other
+vignettes, we'll walk through some basic signal processing tasks using functions
+provided in the `epiprocess` package; of course, we can also write custom code
+for other downstream uses, like plotting, which is pretty easy to do `ggplot2`.
```{r, message = FALSE, warning = FALSE}
library(ggplot2)
@@ -138,7 +138,7 @@ x <- outbreaks::sars_canada_2003 %>%
mutate(geo_value = "ca") %>%
rename(time_value = date) %>%
select(geo_value, time_value, starts_with("cases")) %>%
- as.epi_tibble(geo_type = "nation")
+ as.epi_df(geo_type = "nation")
head(x)
diff --git a/vignettes/issues.Rmd b/vignettes/issues.Rmd
index 718ea1cb..93ac2f0e 100644
--- a/vignettes/issues.Rmd
+++ b/vignettes/issues.Rmd
@@ -1,8 +1,8 @@
---
-title: 6. Work with issue dates and archive objects
+title: 7. Work with issue dates and archive objects
output: rmarkdown::html_vignette
vignette: >
- %\VignetteIndexEntry{6. Work with issue dates and archive objects}
+ %\VignetteIndexEntry{7. Work with issue dates and archive objects}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
@@ -17,7 +17,7 @@ Basic stuff:
```
x <- as_of(z, issue = "2020-06-15")
```
- * here the object `x` will be an `epi_tibble` object
+ * here the object `x` will be an `epi_df` object
* the function `as_of()` by default returns the latest observation (defined
by unique combo of `geo_value` and `time_value` columns being equal) whose
issue doesn't surpass the specified one
@@ -33,11 +33,11 @@ Slide stuff:
- this allows us to do honest forecaster training. basically, replaces `epival`
(we shouldn't need a separate package ...)
- so `epi_slide()` checks what object it's being given
- * if an `epi_tibble`, it works as it does currently
+ * if an `epi_df`, it works as it does currently
* if an `epi_archive`, then for each working time value, it uses `as_of()` to
only return the data you would have had at that time (and whose time value is
withing `n` trailing time steps of the specified one)
- demo this in the vignette with the hand-built AR forecaster. the `epipred`
package can and still should be built, to contain more fancy forecasters, and
potentially also scoring roles. but the functionality for *running* these will
-all already be built into `epitools`, just as a generalization of `epi_slide()`
\ No newline at end of file
+be already built into `epiprocess`, just as a generalization of `epi_slide()`
\ No newline at end of file
diff --git a/vignettes/outliers.Rmd b/vignettes/outliers.Rmd
index 41625ebd..3834e35f 100644
--- a/vignettes/outliers.Rmd
+++ b/vignettes/outliers.Rmd
@@ -8,14 +8,14 @@ vignette: >
---
This vignette describes functionality for detecting and correcting outliers in
-signals `detect_outliers()` and `correct_outliers()`. These functions are
+signals `detect_outliers()` and `correct_outliers()`. These functions are
designed to be modular and extendable, so that you can provide your own outlier
-detection and correction functions and use them with `epi_tibble` objects. We'll
+detection and correction functions and use them with `epi_df` objects. We'll
work state-level daily reported COVID-19 case counts in Florida and New Jersey.
```{r, message = FALSE, fig.width = 8, fig.height = 7}
library(covidcast)
-library(epitools)
+library(epiprocess)
library(dplyr)
library(tidyr)
library(ggplot2)
@@ -29,7 +29,7 @@ x <- covidcast_signal(data_source = "jhu-csse",
geo_values = c("fl", "nj"),
as_of = "2021-10-28") %>%
select(geo_value, time_value, cases = value) %>%
- as.epi_tibble()
+ as.epi_df()
ggplot(x, aes(x = time_value, y = cases)) +
geom_line() +
@@ -128,20 +128,17 @@ plot_outliers = function(x, var, outliers_col = outlier_info,
names_pattern = "(.+)_(.+)")
# If requested, filter to only combined method
- if (combined_only) {
- x_long <- x_long %>% filter(method == "combined")
- }
-
+ if (combined_only) x_long <- x_long %>% filter(method == "combined")
+
# Start of plot with observed data
p <- ggplot() +
geom_line(data = x, mapping = aes(x = time_value, y = !!var))
# If requested, add bands
- if (bands) {
+ if (bands)
p <- p + geom_ribbon(data = x_long,
aes(x = time_value, ymin = lower, ymax = upper,
color = method), fill = NA)
- }
# If requested, add points
if (points) {
@@ -152,9 +149,8 @@ plot_outliers = function(x, var, outliers_col = outlier_info,
}
# If requested, add faceting
- if (!is.null(facet_vars)) {
+ if (!is.null(facet_vars))
p <- p + facet_wrap(facet_vars, nrow = nrow, ncol = ncol, scales = scales)
- }
p
}
diff --git a/vignettes/pct_change.Rmd b/vignettes/pct_change.Rmd
index a5165a30..b53fb853 100644
--- a/vignettes/pct_change.Rmd
+++ b/vignettes/pct_change.Rmd
@@ -9,12 +9,12 @@ vignette: >
A basic way of assessing growth in a signal is to look at its percentage change
over two neighboring time windows. We investigate this in the current vignette,
-using the `pct_change()` function in the `epitools` package. As before, we'll
+using the `pct_change()` function in the `epiprocess` package. As before, we'll
look at state-level daily reported COVID-19 cases.
```{r, message = FALSE}
library(covidcast)
-library(epitools)
+library(epiprocess)
library(dplyr)
x <- covidcast_signal(data_source = "jhu-csse",
@@ -23,21 +23,21 @@ x <- covidcast_signal(data_source = "jhu-csse",
end_day = "2021-05-31",
geo_type = "state",
geo_values = "fl") %>%
- as.epi_tibble() %>%
+ as.epi_df() %>%
rename(cases = value) %>%
select(geo_value, time_value, cases)
```
## Percentage change
-The `pct_change()` function operates on a `epi_tibble` object, and aside from
-this, takes two main arguments: `var`, the variable whose percentage change
-values are to be computed; and `n`, the length of the local window (here, in
-days) to use to compute the percentage change values. For example, if `n = 10`,
-then the percentage change on November 10 is computed as 100 * (B - A) / A,
-where A is the sum of the signal in between November 6 and 10, and A is the sum
-in between November 1 and 5. The default is `n = 14` (here, corresponding to the
-percentage change between adjacent weeks).
+The `pct_change()` function operates on a `epi_df` object, and aside from this,
+takes two main arguments: `var`, the variable whose percentage change values are
+to be computed; and `n`, the length of the local window (here, in days) to use
+to compute the percentage change values. For example, if `n = 10`, then the
+percentage change on November 10 is computed as 100 * (B - A) / A, where A is
+the sum of the signal in between November 6 and 10, and A is the sum in between
+November 1 and 5. The default is `n = 14` (here, corresponding to the percentage
+change between adjacent weeks).
```{r}
x <- pct_change(x, cases, n = 14)
diff --git a/vignettes/slide.Rmd b/vignettes/slide.Rmd
index d949a7f0..46b3c44b 100644
--- a/vignettes/slide.Rmd
+++ b/vignettes/slide.Rmd
@@ -7,31 +7,31 @@ vignette: >
%\VignetteEncoding{UTF-8}
---
-One of the most central tools in the `epitools` package is `epi_slide()`, which
-is based on the family of functions provided by the
+One of the most central tools in the `epiprocess` package is `epi_slide()`,
+which is based on the family of functions provided by the
[`slider`](https://cran.r-project.org/web/packages/slider/) package. In
-`epitools`, to "slide" means to apply a computation---represented as a function
-or formula---over a running window of `n` time steps. Several other functions
-in the `epitools` package, such as `pct_change()` and `estimate_deriv()`, use
-`epi_slide()` as their workhorse. Suitable groupings can always be achieved by
-a preliminary call to `group_by()`.
+`epiprocess`, to "slide" means to apply a computation---represented as a
+function or formula---over a running window of `n` time steps. Several other
+functions in the `epiprocess` package, such as `pct_change()` and
+`estimate_deriv()`, use `epi_slide()` as their workhorse. Suitable groupings can
+always be achieved by a preliminary call to `group_by()`.
By default, the meaning of one time step is inferred from the `time_type` of the
-`epi_tibble` object under consideration. If `time_type` is "day", then one time
-step is one day. The time step can also be specified manually in the call to
+`epi_df` object under consideration. If `time_type` is "day", then one time step
+is one day. The time step can also be specified manually in the call to
`epi_slide()`; you can read the cocumentation for more details. Furthermore, the
-alignment of the running window used in `epi_slide()` can either be "trailing"
+alignment of the running window used in `epi_slide()` can either be "trailing"
or "centered"; the default is "trailing", and is what we use in this and most
other vignettes, without specifying otherwise.
-Similar to the getting started guide, we'll fetch daily reported COVID-19 cases
+Similar to the getting started guide, we'll fetch daily reported COVID-19 cases
for four U.S. states (note: new, not cumulative cases) using the
-[`covidcast`](https://cmu-delphi.github.io/covidcast/covidcastR/) package, and
-then convert this to `epi_tibble` format.
+[`covidcast`](https://cmu-delphi.github.io/covidcast/covidcastR/) package, and
+then convert this to `epi_df` format.
```{r, message = FALSE}
library(covidcast)
-library(epitools)
+library(epiprocess)
library(dplyr)
x <- covidcast_signal(data_source = "jhu-csse",
@@ -40,7 +40,7 @@ x <- covidcast_signal(data_source = "jhu-csse",
end_day = "2021-05-31",
geo_type = "state",
geo_values = c("ca", "fl", "ny", "tx")) %>%
- as.epi_tibble() %>%
+ as.epi_df() %>%
rename(cases = value) %>%
select(geo_value, time_value, cases)
```
@@ -59,14 +59,14 @@ x %>%
```
The formula specified via `slide_fun` has access to all columns present in the
-original `epi_tibble` object (and must refer to them with the prefix `.x$`).
-Here he function `Mean()` is just a simple wrapper around `mean()` that omits
-`NA` values by default (provided by the `epitools` package).
+original `epi_df` object (and must refer to them with the prefix `.x$`). Here
+he function `Mean()` is just a simple wrapper around `mean()` that omits `NA`
+values by default (provided by the `epiprocess` package).
-The function `epi_slide()` returns an `epi_tibble` object with a new column
-appended that contains the results (from sliding), named `slide_value` as the
-default. We can of course change this post hoc, or we can instead specify a new
-name up front using the `new_col_name` argument:
+The function `epi_slide()` returns an `epi_df` object with a new column appended
+that contains the results (from sliding), named `slide_value` as the default. We
+can of course change this post hoc, or we can instead specify a new name up
+front using the `new_col_name` argument:
```{r}
x <- x %>%
@@ -111,20 +111,20 @@ head(x, 10)
## Building and running a local forecaster
As a more complicated example, we create a forecaster based on a local (in time)
-autoregression or AR model. AR models can be fit in numerous ways (using base R
+autoregression or AR model. AR models can be fit in numerous ways (using base R
functions and various packages), but here we define it "by hand" both because it
-provides a more advanced example of sliding a function over an `epi_tibble`
-object, and because it allows us to be a bit more flexible in defining a
-*probabilistic* forecaster: one that outputs not just a point prediction, but
-a notion of uncertainty around this. In particular, our forecaster will
-output a point prediction along with an 90\% uncertainty band, represented by
-a predictive quantiles at the 5\% and 95\% levels (lower and upper endpoints of
-the uncertainty band).
-
-The function defined below, `prob_ar()`, is our probabilistic AR forecaster. The
-`lags`argument indicates which lags to use in the model, and `ahead` indicates
+provides a more advanced example of sliding a function over an `epi_df` object,
+and because it allows us to be a bit more flexible in defining a *probabilistic*
+forecaster: one that outputs not just a point prediction, but a notion of
+uncertainty around this. In particular, our forecaster will output a point
+prediction along with an 90\% uncertainty band, represented by a predictive
+quantiles at the 5\% and 95\% levels (lower and upper endpoints of the
+uncertainty band).
+
+The function defined below, `prob_ar()`, is our probabilistic AR forecaster. The
+`lags`argument indicates which lags to use in the model, and `ahead` indicates
how far ahead in the future to make forecasts (both are encoded in terms of the
-units of the `time_value` column; so, days, in the working `epi_tibble` being
+units of the `time_value` column; so, days, in the working `epi_df` being
considered in this vignette).
```{r}
@@ -155,9 +155,9 @@ prob_ar <- function(v, lags = c(0, 7, 14), ahead = 7, min_train_window = 20,
}
```
-Now we go ahead and slide this AR forecaster over the working `epi_tibble` of
-COVID-19 cases. Note that we actually model the `cases_7dav` column, to operate
-on the scale of smoothed COVID-19 cases. (This is clearly equivalent, up to a
+Now we go ahead and slide this AR forecaster over the working `epi_df` of
+COVID-19 cases. Note that we actually model the `cases_7dav` column, to operate
+on the scale of smoothed COVID-19 cases. (This is clearly equivalent, up to a
constant, to modeling weekly sums of COVID-19 cases.)
```{r}
