Material Management

Description

A set of functions, classes and methods for performing ABC and ABC/XYZ analyses, identifying overperforming, underperforming and constantly performing items, and plotting, analyzing as well as predicting the temporal development of items.

Details

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Class ABCXYZComparison

Description

This S4 class represents the result of a comparison of two ABC/XYZ analysis results.

Slots

data

(data.frame) The comparison result as data.frame.

type

(character) The type of the analysis that has been performed. This is either 'abc' or 'abcxyz'.

valueDiff

(numeric) The difference between the value of an item in ABC/XYZ analysis A and the value of the same item in ABC/XYZ analysis B that is required to consider the item in the comparison.

xyzCoefficientDiff

(numeric) The difference between the xyz coefficient of an item in ABC/XYZ analysis A and the xyz coefficient of the same item in ABC/XYZ analysis B that is required to consider the item in the comparison.

unequalABC

(logical) If TRUE only items are returned, where the ABC-Classes are different. If FALSE only items are returned, where the ABC-Classes are equal. If NA, no further restriction takes place based on the column ABC.

unequalXYZ

(logical) If TRUE only items are returned, where the XYZ-Classes are different. If FALSE only items are returned, where the XYZ-Classes are equal. If NA, no further restriction takes place based on the column XYZ.

Objects from the Class

Objects can be created by calling the function compare function. This S4 class represents the result of a comparison of two ABC/XYZ analysis results.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Examples

data("Amount")
data1 = Amount[sample(1:nrow(Amount), 1000),]
data2 = Amount[sample(1:nrow(Amount), 1000),]
abcxyzData1 = computeABCXYZAnalysis(data1, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
abcxyzData2 = computeABCXYZAnalysis(data2, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
comparison = compare(abcxyzData1, abcxyzData2)
comparison

Class ABCXYZData

Description

This S4 class represents the result of an ABC/XYZ analysis.

Slots

data

(data.frame) The result table of an ABC/XYZ analysis.

type

(character) The type of the analysis that has been performed. This is either 'abc' or 'abcxyz'.

value

(character) The name of the value column in the result table.

item

(character) Vector of the names of the item columns in the result table.

Objects from the Class

Objects can be created by calling the function computeABCXYZ. This S4 class represents the result of an ABC/XYZ analysis.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Examples

data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")
abcResult

Amount data

Description

A dataset containing 23 items and their amounts over 3 years of data.

Usage

Amount

Format

A data frame with 10,000 rows and 9 variables:

date

Date in format yyyy-mm-dd

week

Date in format yyyy-'W'ww

month

Date in format yyyy-mm

quarter

Date in format yyyy-'Q'q

year

Date in format yyyy

item

Item ID

itemgroup

Item group ID

amount

Item amount

value

Item value

Source

anonymized real data

Class Forecast

Description

This S4 class represents the result of forecast using function predictValue.

Slots

data

(data.frame) Data frame including the predicted data and optionally the training data.

models

(list) List of fitted ARIMA models.

value

(character) Name of the value column.

item

(character) Name of the item column.

items

(character) IDs or Names of the items.

Objects from the Class

Objects can be created by calling the function predictValue. This S4 class represents the result of a forecast.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Examples

data("Amount")
prediction = predictValue(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    timeUnitsAhead = 3)
prediction

Stock data

Description

A dataset containing 10 items and their stocks over 3 years of data.

Usage

Stocks

Format

A data frame with 1,610 rows and 5 variables:

date

Date in format yyyy-mm-dd

item

Item ID

stock

Item stock value

minstock

Minimum stock per item

reorderlevel

Stock threshold for triggering item reorders

Source

anonymized real data

Performs a temporal aggregation of a data frame

Description

Aggregates a data frame based on a timestamp column to days, weeks, months, quarters, years or total.

Usage

aggregateData(
  data,
  value = NULL,
  item,
  timestamp,
  temporalAggregation = c("day", "week", "month", "quarter", "year", "total"),
  fiscal = 1,
  aggregationFun = sum
)

Arguments

Value

Returns a data frame with the aggregated data with the columns of item, timestamp and sum, which is the sum of the value column.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

expandData

Examples

data('Amount')
aggregatedData = aggregateData(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date",
    temporalAggregation = "quarter")

Compares two S4 objects

Description

Compares two S4 objects.

Usage

compare(object1, object2, ...)

Arguments

Value

Comparison result.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

compare

Examples

data("Amount")
data1 = Amount[sample(1:nrow(Amount), 1000),]
data2 = Amount[sample(1:nrow(Amount), 1000),]
abcxyzData1 = computeABCXYZAnalysis(data1, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
abcxyzData2 = computeABCXYZAnalysis(data2, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
comparison = compare(abcxyzData1, abcxyzData2)

Compares the results of two ABC/XYZ analyses

Description

Compares the class assignments of two ABC- or two ABC/XYZ analyses.

Usage

## S4 method for signature 'ABCXYZData,ABCXYZData'
compare(
  object1,
  object2,
  valueDiff = NA,
  xyzCoefficientDiff = NA,
  unequalABC = NA,
  unequalXYZ = NA
)

Arguments

Value

An ABCYXZComparison object.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

ABCXYZComparison

Examples

data("Amount")
data1 = Amount[sample(1:nrow(Amount), 1000),]
data2 = Amount[sample(1:nrow(Amount), 1000),]
abcxyzData1 = computeABCXYZAnalysis(data1, value = "value", item = "item", timestamp = "date", 
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
abcxyzData2 = computeABCXYZAnalysis(data2, value = "value", item = "item", timestamp = "date", 
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
comparison = compare(abcxyzData1, abcxyzData2)

Performs an ABC/XYZ analysis

Description

Divides a given data frame into 3 classes, A, B, C, according to the value of one column (e.g., revenue).

Usage

computeABCXYZAnalysis(
  data,
  value,
  item,
  timestamp,
  temporalAggregation = c("day", "week", "month", "quarter", "year"),
  AB = 80,
  BC = 95,
  XY = NA,
  YZ = NA,
  ignoreZeros = FALSE
)

Arguments

Value

Returns an ABCXYZData object. Only positive values are displayed

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

ABCXYZData summary

Examples

# ABC Analysis
data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")

# ABC/XYZ Analysis
data("Amount")
abcxyzResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    XY = 0.3, YZ = 0.5)

Select constant items

Description

Selects items with a constant value for a specified time period.

Usage

computeConstants(
  data,
  value,
  group,
  timestamp,
  timestampFormat = c("day", "week", "month", "quater", "year"),
  currentTime,
  thresholdTime = 7,
  use_latest = FALSE
)

Arguments

Value

Returns a data frame listing all constant items, the date since when the stock is constant and the value of the stock since this time.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

computeUnderperformer computeOverperformer

Examples

data("Stocks")
constants = computeConstants(data=Stocks,
                             value = "stock",
                             group = "item",
                             timestamp = "date",
                             timestampFormat = "day",
                             currentTime = "2019-07-27",
                             thresholdTime = 7,
                             use_latest = FALSE)

Select overperforming items

Description

Selects items with a value higher than a given threshold for a specified time period.

Usage

computeOverperformer(
  data,
  value,
  group,
  timestamp,
  timestampFormat = c("day", "week", "month", "quater", "year"),
  currentTime,
  thresholdValue = 0,
  thresholdTime = 90,
  use_latest = FALSE
)

Arguments

Value

Returns a data frame listing all overperforming items, the date their stock was the last time under the threshold (lastunder), the duration in days since the stock is over the threshold (toolowindays), the average difference between the stock and the threshold (meandiff) and the count of switched between over- and underperformance (moves).

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

computeUnderperformer computeConstants

Examples

data("Stocks")
overperformer = computeOverperformer(data = Stocks,
                     value = "stock",
                     group = "item",
                     timestamp = "date",
                     timestampFormat = "day",
                     currentTime = "2019-07-27",
                     thresholdValue = "reorderlevel",
                     thresholdTime = 0,
                     use_latest = FALSE)

Select underperforming items

Description

Selects items with a value lower than a given threshold for a specified time period.

Usage

computeUnderperformer(
  data,
  value,
  group,
  timestamp,
  timestampFormat = c("day", "week", "month", "quater", "year"),
  currentTime,
  thresholdValue = 0,
  thresholdTime = 90,
  use_latest = FALSE
)

Arguments

Value

Returns a data frame listing all underperforming items, the date their stock was the last time over the threshold (lastover), the duration in days since the stock is under the threshold (toolowindays), the average difference between the stock and the threshold (meandiff) and the count of switched between over- and underperformance (moves).

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

computeOverperformer computeConstants

Examples

data("Stocks")
underperformer = computeUnderperformer(data=Stocks,
                                       value = "stock",
                                       group = "item",
                                       timestamp = "date",
                                       timestampFormat = "day",
                                       currentTime = "2019-07-27",
                                       thresholdValue = "minstock",
                                       thresholdTime = 90,
                                       use_latest = FALSE)

Detects items whose value (stock, demand, etc.) has changed over time

Description

Detects items whose value (stock, demand, etc.) has changed over time in contrast to other items. This analysis is based on the Macnaughton-Smith et al. clustering algorithm.

Usage

detectTimeVariations(
  data,
  value,
  item,
  timestamp,
  temporalAggregation = c("day", "week", "month", "quarter", "year"),
  aggregationFun = sum,
  preProcess = NA,
  recentTimePeriods = 5
)

Arguments

Value

Returns a data frame showing to which cluster each item belongs based on all value and based on the recent values as well as whether the item has switched the cluster.

References

Macnaughton-Smith, P., Williams, W.T., Dale, M.B., Mockett, L.G. (1964) "Dissimilarity Analysis: a new Technique of Hierarchical Sub-division", Nature, 202, 1034–1035.

Examples

data("Amount")
timeVariations = detectTimeVariations(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week")

Expands a temporal data frame

Description

Expands a temporal data frame and fills values for missing dates.

Usage

expandData(
  data,
  expand,
  expandTo = c("all", "event"),
  valueColumns,
  latest_values = F,
  valueLevels = NA,
  timestamp,
  timestampFormat = c("day", "week", "month", "quarter", "year"),
  keepData = T
)

Arguments

Value

Returns the expanded data frame.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

aggregateData

Examples

data("Amount")
expandedItems = expandData(Amount,
    expand = c("item", "itemgroup"),
    expandTo = "all",
    valueColumns = c("amount", "value"),
    latest_values = TRUE,
    valueLevels = c(0, 0),
    timestamp = "date",
    timestampFormat = "day")

Launches a demo app

Description

Launches a shiny app that demonstrates how to use the functions provides by package matman.

Usage

matmanDemo()

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Examples

## Not run: matmanDemo()

Plots the result of an ABC/XYZ analysis

Description

Plots a graph that shows what percentage of items is responsible for what amount of value.

Usage

## S4 method for signature 'ABCXYZData,ANY'
plot(
  x,
  plot_engine = c("graphics", "plotly"),
  title = "",
  xlab = "",
  ylab = "",
  top5lab = NA,
  color = list(itemColor = "blue", top5Color = "black", aColor = "green", bColor =
    "orange", cColor = "red"),
  item = NA,
  ...
)

Arguments

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

computeABCXYZAnalysis ABCXYZData

Examples

data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")
plot(abcResult,
     plot_engine = "graphics",
     title = "ABC Analysis",
     xlab = "Items",
     ylab = "Demand")

Plots the development of the values

Description

Plots a bar chart that shows the sum of the value column for a certain time interval.

Usage

plotValueSeries(
  data,
  item,
  item_id,
  value,
  timestamp,
  temporalAggregation = c("day", "week", "month", "quarter", "year"),
  expand = TRUE,
  withTrendLine = TRUE,
  windowLength = 5,
  trendLineType = "s"
)

Arguments

Value

A plotly bar chart, that shows the development of the value column.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

Examples

data("Amount")
plotValueSeries(Amount,
                item = "item",
                item_id = "45186",
                value = "amount",
                timestamp = "date",
                temporalAggregation = "week",
                withTrendLine = TRUE,
                windowLength = 10,
                trendLineType = "w")

Predicts the value for items

Description

Predicts the value for items based on previous values. Previous values can be aggregated to value per day, week, month, quarter or year. An ARIMA model is estimated for each item based on the function forecast:auto.arima. The best model is selected and used for prediction. Note that only models without drift term will be considered in order to ensure consistent predictions.

Usage

predictValue(
  data,
  value,
  item,
  timestamp,
  temporalAggregation = c("day", "week", "month", "quarter", "year"),
  aggregationFun = sum,
  timeUnitsAhead = 1,
  digits = 3,
  expand = F,
  keepPreviousData = F,
  level = 0.95,
  ...
)

Arguments

Value

Returns a Forecast object.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

auto.arima Forecast

Examples

# Simple Example
data("Amount")
prediction = predictValue(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    timeUnitsAhead = 3)
prediction

# More Sophisticated Example
data("Amount")
prediction = predictValue(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    aggregationFun = mean,
    timeUnitsAhead = 5,
    digits = 4,
    keepPreviousData = TRUE,
    level = 0.9,
    trace = TRUE)
prediction

Shows an ABCXYZComparison object

Description

Shows an ABCXYZComparison object as a table consisting of the absolute and relative amount of each item, the cumulative relative amount and the ABC-class for both ABCXYZData objects. It furthermore shows the ABC comparison of the two objects. If XY and YZ parameters have been specified for computing the ABCXYZData object, the table also includes a column for the XYZ coefficient, the XYZ-class, the ABC/XYZ-class and the XYZ comparison.

Usage

## S4 method for signature 'ABCXYZComparison'
show(object)

Arguments

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

ABCXYZComparison compare

Examples

data("Amount")
data1 = Amount[sample(1:nrow(Amount), 1000),]
data2 = Amount[sample(1:nrow(Amount), 1000),]
abcxyzData1 = computeABCXYZAnalysis(data1, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
abcxyzData2 = computeABCXYZAnalysis(data2, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
comparison = compare(abcxyzData1, abcxyzData2)
comparison

Shows an ABCXYZData object

Description

Shows the ABCXYZData object as a table consisting of the absolute and relative amount of each item, the cumulative relative amount and the ABC-class. If XY and YZ parameters have been specified for computing the ABCXYZData object, the table also includes a column for the XYZ coefficient, the XYZ- class and the ABC/XYZ-class.

Usage

## S4 method for signature 'ABCXYZData'
show(object)

Arguments

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

ABCXYZData computeABCXYZAnalysis

Examples

data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")
abcResult

Shows a Forecast object

Description

Shows the predicted data of a Forecast object. If the Forecast object was created using keepPreviousData = TRUE, also the training data are shown

Usage

## S4 method for signature 'Forecast'
show(object)

Arguments

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

Forecast

Examples

data("Amount")
prediction = predictValue(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    timeUnitsAhead = 3)
prediction

Summarizes an S4 object

Description

Summarizes an S4 object.

Usage

summary(object, ...)

Arguments

Value

Summary.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

summary summary summary

Examples

data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")
summary(abcResult)

Prints the summary of the comparison of two ABC/XYZ analyses

Description

Summarizes the differences between two ABCXYZData objects.

Usage

## S4 method for signature 'ABCXYZComparison'
summary(object, withMissing = FALSE)

Arguments

Value

A contingency table showing the differences.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

compare ABCXYZComparison

Examples

data("Amount")
data1 = Amount[sample(1:nrow(Amount), 1000),]
data2 = Amount[sample(1:nrow(Amount), 1000),]
abcxyzData1 = computeABCXYZAnalysis(data1, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
abcxyzData2 = computeABCXYZAnalysis(data2, value = "value", item = "item", timestamp = "date",
                                    temporalAggregation = "day", XY = 0.5, YZ = 1)
comparison = compare(abcxyzData1, abcxyzData2)
summary(comparison)

Prints the result summary of an ABC/XYZ analysis

Description

Summarizes the items count and value sum grouped by the different ABC- or ABC/XYZ-Classes.

Usage

## S4 method for signature 'ABCXYZData'
summary(object, withMissing = FALSE)

Arguments

Value

A data.table with the summarized results.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

computeABCXYZAnalysis ABCXYZData

Examples

# ABC Analysis
data("Amount")
abcResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date")
summary(abcResult)

# ABC/XYZ Analysis
data("Amount")
abcxyzResult = computeABCXYZAnalysis(data = Amount,
    value = "value",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    XY = 0.3, YZ = 0.5)
summary(abcxyzResult)

Prints the summary of a Forecast object

Description

Summarizes the fitted models estimated for predicting item values (e.g., demand, stock).

Usage

## S4 method for signature 'Forecast'
summary(object)

Arguments

Value

A data frame showing a summary of fitted models.

Author(s)

Leon Binder leon.binder@th-deg.de

Bernhard Bauer bernhard.bauer@th-deg.de

Michael Scholz michael.scholz@th-deg.de

See Also

predictValue Forecast

Examples

data("Amount")
prediction = predictValue(data = Amount,
    value = "amount",
    item = "item",
    timestamp = "date",
    temporalAggregation = "week",
    timeUnitsAhead = 3)
summary(prediction)