Sentry Page Protection
Summary
Proc Univariate
1. Proc Univariate is commonly used in Exploratory Analysis.
It computes the 5 sets of statistical measurements such as:
2. There are two approaches to normality test:
3. Use the CLASS statement when you have multiple segments, and use the OUTPUT statement when having to export the analysis results to a data set.
Proc Means
4. Proc Means is similar to Proc Univariate. It generates summary statistics such as
5. Similar to Proc Univariate, use the CLASS statement on Proc Means when analyzing data from multiple segments and the OUTPUT statement when having to export the analysis results to a data set.
6. The _TYPE_ variable from the output data set identifies the analysis results from different classification levels.
Proc Freq
7. Proc Freq computes counting statistics on categorical variables. It works with the TABLE statement as opposed to the VAR statement (from Proc Means/Univariate).
8. The PLOTS option allows you to plot the histogram and cumulative frequency plot, which is an easy way to visualize the data.
9. The WEIGHT statement should be used when the observations have unequal weight.
10. The n-way crosstabulation table provides insights on the relationship between multiple variables.
11. Use the OUT option to export analysis results to a data set.
Proc Univariate
1. Proc Univariate is commonly used in Exploratory Analysis.
It computes the 5 sets of statistical measurements such as:
- Moments
- Basic Statistical Measures
- Tests for Location
- Quantiles
- Extreme Values
2. There are two approaches to normality test:
- Numerical Method
The numeric method looks at the 4 normal testings:- Shapiro-Wilk
- Kolmogorov-Smirnov
- Cramer-von Mises
- Anderson-Darling
- Graphical Approach
The graphical method looks at the 4 distribution plots:- Stem-and-leaf plot (or a horizontal bar chart)
- Box plot
- Normal probability plot
- Histogram
3. Use the CLASS statement when you have multiple segments, and use the OUTPUT statement when having to export the analysis results to a data set.
Proc Means
4. Proc Means is similar to Proc Univariate. It generates summary statistics such as
- Mean
- Standard Deviation
- Minimum, Maximum, Median, Mode
- Confidence Limits, Quantiles
5. Similar to Proc Univariate, use the CLASS statement on Proc Means when analyzing data from multiple segments and the OUTPUT statement when having to export the analysis results to a data set.
6. The _TYPE_ variable from the output data set identifies the analysis results from different classification levels.
Proc Freq
7. Proc Freq computes counting statistics on categorical variables. It works with the TABLE statement as opposed to the VAR statement (from Proc Means/Univariate).
8. The PLOTS option allows you to plot the histogram and cumulative frequency plot, which is an easy way to visualize the data.
9. The WEIGHT statement should be used when the observations have unequal weight.
10. The n-way crosstabulation table provides insights on the relationship between multiple variables.
11. Use the OUT option to export analysis results to a data set.