blockr.ts

Time Series Analysis Blocks for blockr

Specialized time series functionality for blockr using tsbox and dygraphs. Provides comprehensive blocks for loading, transforming, and visualizing time series data as interactive dygraphs.

Features

• 📊 Interactive Dygraphs: All outputs render as interactive time series charts
• 📦 25 Built-in Datasets: Easy access to R's time series datasets
• 🔄 Comprehensive Transformations: Changes, frequency conversion, lag/lead, scaling
• 📈 Advanced Analysis: Forecasting, decomposition, PCA for multivariate series
• 🎨 Professional UI: Consistent design following blockr.ggplot patterns
• 🎯 tsbox Integration: Seamless format conversion and manipulation

Installation

# Install dependencies
install.packages(c("tsbox", "dygraphs", "forecast"))

# Install blockr.core if not already installed
# remotes::install_github("blockr-org/blockr.core")

# Install blockr.ts
devtools::install()

Quick Start

library(blockr.core)
library(blockr.ts)

# Create a pipeline using new_board()
blockr.core::serve(
  new_board(
    blocks = c(
      data = new_ts_dataset_block(dataset = "AirPassengers"),
      transform = new_ts_change_block(method = "pcy")
    ),
    links = c(
      data_to_transform = new_link("data", "transform")
    )
  )
)

Available Blocks

Data Blocks

new_ts_dataset_block()

Access all 25 built-in R time series datasets with an intuitive selector.

blockr.core::serve(
  new_ts_dataset_block(dataset = "EuStockMarkets")
)

Transform Blocks

new_ts_change_block()

Calculate various types of changes: percentage, differences, year-over-year.

blockr.core::serve(
  new_ts_change_block(method = "pcy"),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

Methods:

• pc: Period-on-period percentage change
• pcy: Year-on-year percentage change
• pca: Annualized percentage change
• diff: First differences
• diffy: Year-on-year differences

new_ts_frequency_block()

Convert time series between temporal granularities with smart aggregation.

blockr.core::serve(
  new_ts_frequency_block(to = "year", aggregate = "mean"),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

Features:

• Automatic frequency detection
• Smart target selection (only allows aggregation)
• Multiple aggregation methods: mean, sum, first, last, min, max

new_ts_select_block()

Select specific series from multivariate time series data.

multivariate_data <- tsbox::ts_c(datasets::mdeaths, datasets::fdeaths)

blockr.core::serve(
  new_ts_select_block(series = "mdeaths", multiple = FALSE),
  data = list(data = tsbox::ts_tbl(multivariate_data))
)

new_ts_lag_block()

Shift time series forward (lag) or backward (lead).

blockr.core::serve(
  new_ts_lag_block(by = 12),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

new_ts_span_block()

Filter time series to specific date ranges with an intuitive range slider.

blockr.core::serve(
  new_ts_span_block(start = 1950, end = 1955),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

Analysis Blocks

new_ts_scale_block()

Scale, normalize, or index time series for comparison.

blockr.core::serve(
  new_ts_scale_block(method = "index"),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

Methods:

• normalize: Scale to mean=0, sd=1
• index: Index to base period (100)
• minmax: Scale to [0, 1] range

new_ts_decompose_block()

Extract trend, seasonal, and remainder components.

blockr.core::serve(
  new_ts_decompose_block(component = "seasonal_adjusted"),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

new_ts_forecast_block()

Generate forecasts with confidence intervals.

blockr.core::serve(
  new_ts_forecast_block(horizon = 24),
  data = list(data = tsbox::ts_tbl(datasets::AirPassengers))
)

new_ts_pca_block()

Principal Component Analysis for multivariate time series.

blockr.core::serve(
  new_ts_pca_block(n_components = 2),
  data = list(data = tsbox::ts_tbl(datasets::EuStockMarkets))
)

new_ts_to_df_block()

Convert time series data to regular data frame format, removing the dygraph visualization. Useful for integrating with non-TS blocks or exporting data.

blockr.core::serve(
  new_ts_to_df_block(format = "long"),
  data = list(data = tsbox::ts_tbl(tsbox::ts_c(datasets::mdeaths, datasets::fdeaths)))
)

blockr.core::serve(
  new_ts_to_df_block(format = "wide"),
  data = list(data = tsbox::ts_tbl(tsbox::ts_c(datasets::mdeaths, datasets::fdeaths)))
)

new_ts_from_df_block()

Convert wide-format data frames to time series format with dygraph visualization. Automatically detects time columns and converts numeric columns to separate series.

df_wide <- data.frame(
  date = seq(as.Date("2020-01-01"), by = "month", length.out = 24),
  sales = rnorm(24, 100, 10),
  revenue = rnorm(24, 1000, 100),
  costs = rnorm(24, 500, 50)
)

blockr.core::serve(
  new_ts_from_df_block(),
  data = list(data = df_wide)
)

Advanced DAG Board Pipeline

For complex analyses with multiple branches, use the DAG board:

library(blockr.core)
library(blockr.ui)
library(blockr.dplyr)
library(blockr.ggplot)
library(blockr.ai)

# Create comprehensive analysis board
ts_board <- blockr.ui::new_dag_board(
  blocks = c(
    # Data sources
    air = new_ts_dataset_block(dataset = "AirPassengers"),
    stocks = new_ts_dataset_block(dataset = "EuStockMarkets"),

    # AirPassengers branch
    air_decomp = new_ts_decompose_block(component = "seasonal_adjusted"),
    air_change = new_ts_change_block(method = "pcy"),
    air_forecast = new_ts_forecast_block(horizon = 24),

    # Stocks branch
    stock_select = new_ts_select_block(series = c("DAX", "FTSE")),
    stock_scale = new_ts_scale_block(method = "index"),
    stock_pca = new_ts_pca_block(n_components = 2)
  ),

  links = c(
    # Connect AirPassengers pipeline
    new_link("air", "air_decomp", "data"),
    new_link("air_decomp", "air_change", "data"),
    new_link("air_change", "air_forecast", "data"),

    # Connect stocks pipeline
    new_link("stocks", "stock_select", "data"),
    new_link("stock_select", "stock_scale", "data"),
    new_link("stock_scale", "stock_pca", "data")
  )
)

# Serve the board
blockr.core::serve(ts_board)

See Also

• blockr.core - Core blockr framework
• blockr.ggplot - ggplot2 visualization blocks
• tsbox - Time series toolbox
• dygraphs - Interactive time series charts