Automated Capability Analysis

SAS macros to automate / simplify the process of capability analysis performed in SAS.

SAS Code Documentation

This SAS code defines several functions for working with significant figures and reformatting numbers. Below is a summary of each function, the parameters they accept, and the output they provide:

rndgt(x, sf)

• Description: This function rounds a number x to the specified number of significant figures (sf).

• Parameters:

  • x: The number to be rounded.
  • sf: The desired number of significant figures.

• Output: The rounded number with sf significant figures.

extsf(x)

• Description: This function extracts significant digits from a number x.

• Parameters:

  • x: The number from which significant digits are to be extracted.

• Output: A character string representing the significant digits of x.

sigFig(x, sf)

• Description: This function displays a number x with the specified number of significant figures (sf) as a character string.

• Parameters:

  • x: The number to be displayed.
  • sf: The desired number of significant figures.

• Output: A character string representation of x with sf significant figures. Trailing zeros are added if necessary.

re_format(x)

• Description: This function reformats a number x by either returning the integer part of x if it has three or more digits or applying the sigFig function with three significant figures to x if it has fewer than three digits.

• Parameters:

  • x: The number to be reformatted.

• Output: The reformatted number.

re_format_dataset(Data_In, Var)

• Description: This macro reformats a dataset by applying formatting conditions to a specified variable.

• Parameters:

  • Data_In: The name of the input dataset to be reformatted.
  • Var: The name of the variable within the dataset to be reformatted.

• Reformatting Logic:

  • If the value of Var is greater than 0:
    • If Var is greater than or equal to 100, it is rounded down to the nearest integer using the floor function.
    • If Var is less than 100, it is reformatted to have three significant figures using the SigFig function.
  • If Var is exactly 0, it remains unchanged as 0.
  • If Var is less than 0:
    • If Var is less than or equal to -100, it is rounded up to the nearest integer using the ceil function.
    • If Var is greater than -100, it is reformatted to have three significant figures using the SigFig function while preserving its negative sign.

• Output: The dataset specified by Data_In is modified in place with the variable Var reformatted according to the specified conditions. The Var variable is also formatted using the BEST. format.

• Usage: To use this macro, call it by passing the name of the input dataset (Data_In) and the name of the variable to be reformatted (Var) as arguments.

  Example Usage:

  %re_format_dataset(MyData, MyVariable);

Get_data(Data_In, Line_Loc, Var, Data_Out_Var)

• Description: This macro retrieves the value from a specified cell in a dataset and assigns it to a macro variable.

• Parameters:

  • Data_In: The name of the dataset from which the value will be retrieved.
  • Line_Loc: The row index (observation number) of the cell from which the value will be retrieved.
  • Var: The column name (variable) of the cell from which the value will be retrieved.
  • Data_Out_Var: The name of the macro variable to which the retrieved value will be assigned.

• Functionality:

  • The macro initializes a global macro variable named &Data_Out_Var to hold the retrieved data.
  • It then reads through the dataset specified by Data_In and, when it reaches the row specified by Line_Loc, it assigns the value of the variable Var from that row to the macro variable &Data_Out_Var.
  • The assignment is done using the symput function, which assigns the value of Var to the macro variable named in Data_Out_Var.

• Output: The macro variable specified by Data_Out_Var will contain the value retrieved from the specified cell in the dataset.

• Usage: To use this macro, call it by passing the name of the input dataset (Data_In), the row index (Line_Loc), the column name (Var), and the name of the macro variable to store the data (Data_Out_Var).

  Example Usage:

  %Get_data(MyDataset, 3, MyVariable, OutputVariable);

Check_Condition(Var, Out)

• Description: This macro converts a numeric variable (Var) into a TRUE or FALSE condition and assigns the result to a macro variable (Out) based on a specified threshold.

• Parameters:

  • Var: The numeric variable to be evaluated.
  • Out: The name of the macro variable where the TRUE or FALSE condition will be stored.

• Functionality:

  • The macro initializes a global macro variable named &Out to hold the TRUE or FALSE condition.
  • It evaluates the value of the variable Var against the threshold of 1.33.
  • If Var is less than 1.33, it assigns the string "TRUE" to the macro variable &Out. Otherwise, it assigns "FALSE."

• Output: The macro variable specified by Out will contain either "TRUE" or "FALSE" based on whether the condition with the threshold was met.

• Usage: To use this macro, call it by passing the numeric variable (Var) that you want to evaluate and the name of the macro variable (Out) where the TRUE or FALSE condition will be stored.

  Example Usage:

  %Check_Condition(MyNumericVar, MyCondition);

get_sentence(Data_In, Sentence_Loc, Data_Out_Var)

• Description: This macro filters a dataset (Data_In) to retrieve a specific sentence identified by Sentence_Loc and stores it in a macro variable (Data_Out_Var).

• Parameters:

  • Data_In: The name of the dataset containing sentences.
  • Sentence_Loc: The location or identifier of the target sentence to retrieve.
  • Data_Out_Var: The name of the macro variable where the retrieved sentence will be stored.

• Functionality:

  • The macro creates a temporary dataset called temp_sentence by filtering the rows in Data_In where the variable Sentence matches the value of Sentence_Loc.
  • It then checks the number of observations (NOBS) in the temp_sentence dataset and initializes the macro variable &Data_Out_Var to an empty string.
  • If there are observations in the temp_sentence dataset (i.e., NOBS > 0), the macro retrieves the Text column from the first observation and stores it in the macro variable &Data_Out_Var.

• Output: The macro variable specified by Data_Out_Var will contain the retrieved sentence, or it will be an empty string if the specified sentence is not found in the dataset.

• Usage: To use this macro, call it by passing the name of the input dataset (Data_In), the identifier of the target sentence (Sentence_Loc), and the name of the macro variable (Data_Out_Var) where the retrieved sentence will be stored.

  Example Usage:

  %get_sentence(MyDataset, "Identifier123", MySentence);

Histogram_Plot(Data_In, LSL, USL, Var_Con, Title_Label, Footnote, Image_Loc, Image_Name, DPI)

• Description: This macro generates a histogram plot for a variable (Var_Con) in a dataset (Data_In) and includes control limits (LSL and USL) as reference lines. It allows customization of various plot settings.

• Parameters:

  • Data_In: The name of the dataset containing the variable to be plotted.
  • LSL: The Lower Specification Limit (control limit) for the variable.
  • USL: The Upper Specification Limit (control limit) for the variable.
  • Var_Con: The name of the variable to be plotted in the histogram.
  • Title_Label: The title to be displayed on the histogram plot.
  • Footnote: The footnote to be displayed on the histogram plot.
  • Image_Loc: The location where the generated plot image will be saved.
  • Image_Name: The name to be assigned to the generated plot image.
  • DPI: The resolution (Dots Per Inch) for the generated image.

• Functionality:

  • The macro sets ODS (Output Delivery System) options to control the output format and image quality.
  • It uses PROC MEANS to calculate summary statistics for the specified variable (Var_Con) and stores them in a table called descriptive_table.
  • The macro extracts the minimum (Min) and maximum (Max) values from the descriptive_table and assigns them to macro variables.
  • It determines suitable lower (LOWER) and upper (UPPER) limits for the x-axis based on the specified control limits (LSL and USL) and the minimum and maximum values.
  • Using PROC SGPLOT, the macro generates a histogram plot with the specified variable, fills the histogram bars with a specified color, adds reference lines for control limits, and customizes axis labels and settings.
  • The generated plot is saved as an HTML image with the specified resolution and filename.

• Output: The macro generates a histogram plot for the specified variable with control limits and saves it as an image in the specified location.

• Usage: To use this macro, call it by passing the necessary parameters, including the dataset name, variable name, control limits, plot title, and other customization options.

  Example Usage:

  %Histogram_Plot(MyDataset, 10, 30, MyVariable, "Histogram of MyVariable", "Footnote Text", "C:\Output\Images\", MyHistogram, 300);

Histogram_Plot(Data_In, LSL, USL, Var_Con, Title_Label, Footnote, Image_Loc, Image_Name, DPI)

• Description: This macro generates a histogram plot for a variable (Var_Con) in a dataset (Data_In) and includes control limits (LSL and USL) as reference lines. It allows customization of various plot settings.

• Parameters:

  • Data_In: The name of the dataset containing the variable to be plotted.
  • LSL: The Lower Specification Limit (control limit) for the variable.
  • USL: The Upper Specification Limit (control limit) for the variable.
  • Var_Con: The name of the variable to be plotted in the histogram.
  • Title_Label: The title to be displayed on the histogram plot.
  • Footnote: The footnote to be displayed on the histogram plot.
  • Image_Loc: The location where the generated plot image will be saved.
  • Image_Name: The name to be assigned to the generated plot image.
  • DPI: The resolution (Dots Per Inch) for the generated image.

• Functionality:

  • The macro sets ODS (Output Delivery System) options to control the output format and image quality.
  • It uses PROC MEANS to calculate summary statistics for the specified variable (Var_Con) and stores them in a table called descriptive_table.
  • The macro extracts the minimum (Min) and maximum (Max) values from the descriptive_table and assigns them to macro variables.
  • It determines suitable lower (LOWER) and upper (UPPER) limits for the x-axis based on the specified control limits (LSL and USL) and the minimum and maximum values.
  • Using PROC SGPLOT, the macro generates a histogram plot with the specified variable, fills the histogram bars with a specified color, adds reference lines for control limits, and customizes axis labels and settings.
  • The generated plot is saved as an HTML image with the specified resolution and filename.

• Output: The macro generates a histogram plot for the specified variable with control limits and saves it as an image in the specified location.

• Usage: To use this macro, call it by passing the necessary parameters, including the dataset name, variable name, control limits, plot title, and other customization options.

  Example Usage:

  %Histogram_Plot(MyDataset, 10, 30, MyVariable, "Histogram of MyVariable", "Footnote Text", "C:\Output\Images\", MyHistogram, 300);

Generate_difference(Ordered_Difference, p_value, vlabel)

• Description: This macro generates a sentence summarizing the results of an ANOVA analysis, including the p-value and whether it is below or above a significance level (e.g., 0.05).

• Parameters:

  • Ordered_Difference: The name of the dataset containing differences between groups.
  • p_value: The p-value resulting from the ANOVA analysis.
  • vlabel: A label or description of the variable being analyzed.

• Functionality:

  • The macro first handles the p-value formatting. If the p-value is "<.0001," it is converted to "0.0001." Otherwise, it is converted to a numeric value using the INPUT function.
  • It then constructs an initial sentence (Current_State) describing the ANOVA analysis and the variable of interest (vlabel) with a placeholder for the p-value.
  • Based on whether the p-value is less than 0.05, the macro determines whether to state that the p-value is "below" or "above" a significance level.
  • The Ordered_Difference dataset is processed to handle p-value formatting and to create a new variable called Statement that describes the positions being compared in the ANOVA.
  • Depending on the p-value condition, the Current_State sentence is updated to include "below" or "above" significance.

• Output: The macro updates the Current_State sentence to describe the ANOVA results, including whether the p-value is below or above a significance level.

• Usage: To use this macro, call it by passing the necessary parameters, including the dataset name containing differences between groups (Ordered_Difference), the p-value, and a label for the variable being analyzed (vlabel).

  Example Usage:

  %Generate_difference(MyDifferences, 0.0001, "MyVariable");

pp_capability(Data_In, Var_Con, Var_Cat, Image_Loc, Output_Excel, Threshold, LSL, USL, Encode, units = mg/100ml, Main_Title = Capability Results)

• Description: This macro generates a PowerPoint slide prototype for displaying capability analysis results, including histograms, box plots, descriptive statistics, and textual information.

• Parameters:

  • Data_In: The name of the dataset containing the analysis data.
  • Var_Con: The continuous variable of interest.
  • Var_Cat: The categorical variable for grouping.
  • Image_Loc: The location where generated images will be saved.
  • Output_Excel: The name of the Excel file containing analysis results.
  • Threshold: The threshold for significance.
  • LSL: The Lower Specification Limit.
  • USL: The Upper Specification Limit.
  • Encode: A dataset containing encoding for text replacement.
  • units: Units of measurement for the variable (default: "mg/100ml").
  • Main_Title: The main title for the slide (default: "Capability Results").

• Functionality:

  • The macro starts by setting up the output environment for PowerPoint and importing relevant data.
  • It prepares the data for analysis and retrieves necessary values for generating the slide.
  • Descriptive statistics, histograms, and box plots are generated and saved as images.
  • The ANOVA analysis is run, and p-values and differences are retrieved.
  • Text replacement is performed based on encoding provided in the Encode dataset.
  • The slide layout is designed with titles, text, tables, and images, and all elements are placed on the slide as per the prototype design.

• Output: The macro generates a PowerPoint slide prototype with various elements, including images, tables, and text, to display capability analysis results.

• Usage: To use this macro, call it by passing the necessary parameters, including the dataset name, variable names, and other customization options.

  Example Usage:

  %pp_capability(MyData, Measurement, Group, "C:\Output\Images\", "AnalysisResults.xlsx", 0.05, 10, 30, EncodeData, "mg/100ml", "Capability Analysis");