CLASSIFY COMMAND

Whenever you are running ACL, use the Classify Command on every character-based column which is 4 or fewer
characters wide.

Description of the Classify Command:

The Classify Command is one of the many functions within ACL. It basically provides totals for different items in a
column. Examples:

You have a column for STATE and a column for INVOICE AMOUNT. Classify will give you a total INVOICE
AMOUNT per STATE.

You have a column for LOAN TYPE and a column for LOAN PRINCIPAL BALANCE. Classify will give you a total
LOAN PRINCIPAL BALANCE per LOAN TYPE.

You have a column for STORE NUMBER and a column for EXTENDED INVENTORY. Classify will give you a total
INVENTORY per STORE location.

The results look as follows:

TYPE COUNT <-- % % --> Trans Amount

AA 1 0.13% -0.11% -533.59

CN 108 13.99% -1.92% -9025.02
IN 588 76.17% 112.02% 525259.16
PM 71 9.20% -9.66% -45281.38
TR 4 0.52% -0.33% -1538.48

772 100.00% 100.00% 468880.69

In this case is an A/R transaction file. One of the columns is called TYPE which represents transaction type. It is a
character column less than 4 characters wide. The results represent total transaction amount per transaction type. The
following describes the columns:

TYPE is the name of the field being classified. If I wanted to classify on STATE, the column heading would have been
"STATE".

COUNT represents the number of records (rows) per unique item in the field being classified on; in this case, TYPE. There
were 108 transactions with the type being CN (credit notice).

<-- % represents the percentage of records (rows) for that TYPE as compared to the total number of records (rows) in the
file. For type CN, the 108 records is 13.99% of the total number of records, 772.

% --> represents the percentage of the accumulating column total for each TYPE to the total for A/R. That is, the total PM
which is -45281.38 is -9.66% of the total A/R transaction amount, 468,880.69.

To use the classify command in ACL:

1. Identify the column you want to classify on.
2. Identify the column you want to total or accumulate. 3. In ACL, click on the Data menu and then on Classify.
4. In the dialog box, select the field(column) to classify on, and the one to accumulate (total).
5. Click on the button "Output to" and then on "Screen".
6. Click OK twice and you will have output like the example above.

A common error occurs when you try to classify and the field to classify on is numeric. You can use Edit Expressions to
create a new field which is character. Use the STRING() function – call an ACL Resource to get more information.

The Classify Command is like an X-RAY of the client's data/business. You often find things that are unexpected. E.g.,:
- Sales in a particular state greater than expected - Sales in states you can't locate in an atlas
- Loan types you never thought the client had

Page 1
STATISTICS COMMAND
Run the Statistics Command on every numeric and date column in your data file.

Description of the Statistics Command:

The Statistics Command in ACL is a simple click of a button which generates high level information about a numeric
column. For example, the following is an A/R file downloaded into ACL. The file contains all invoices, and credit memos.
The column for the A/R amount is called "amt". The following are the results after running Statistics on the amt
column(field):

Field : amt
Number Total Average

Positive: 57,145 85,511,105.41 1,496.39

Zeros: 0
Negative: 3,772 -3,643,427.79 -965.91
Totals: 60,917 81,867,677.62 1,343.92
Abs Value: 89,154,533.20
Range: 903,563.73

Highest : 808,441.73 799,341.45 726,360.87 474,654.38 464,217.99

Lowest : -95,122.00 -87,276.75 -54,323.22 -50,290.40 -50,000.00

The following describes the output:

- There are three basic columns, Number, Total and Average. Total and Average are self-explanatory. Number represents
the number of records or rows of data (i.e., there were 57,145 invoices with positive amounts, 3,772 credits and a total
number of rows of 60,917).

- Along the left side, Positive means records with positive amounts; Zeros represents zero value amounts; Negative
represents records with negative amounts; Totals is the grand total and should tie to something in the G/L; Abs Value is the
Positive amount plus the Negative; and Range is the distance between the highest individual amount and the lowest.

- At the bottom, there are two rows, Highest and Lowest. These are the five highest amounts and the five lowest amounts.
In this case, the largest invoice for the year was $808,441.73 and the largest credit was $-95,122.00.

To use the Statistics Command in ACL

The easiest way to use the Statistics Command is to:

1. Click the column heading of the amount you want statistics on.
2. Click the Statistics icon

NOTE: The Statistics icon is the bell-shaped-curve with the + sign.

You can also use the menu. Statistics is in the Analyze Menu.

If you do not click the column heading, and therefore do not highlight a column, the Statistics command will provide a list
of numeric and date fields. Highlight one or more and click on [OK].

In that dialog box, you will notice a couple of options as well:

- If you check the box for Std Deviation, the standard deviation will appear in the output under Range.

- You can put something under "IF", like BRANCH="04" or AMT > 50000 (assuming these are valid column names and
values), and the statistics will apply to all records meeting these criteria. Or you could use a filter rather than the "IF" in the
dialog box.

Page 2
Innovative uses of the Statistics Command:

Use the Statistics Command after applying a filter. If you use a filter, you can run many commands while it is in effect. If
you use the "IF" in the Statistics Command dialog box instead of a filter, it only applies to that one time you run the
command.

As a refresher, the filter in ACL is just above your data on the screen. It is the area you can type in, just to the right of the
yellow arrow, red x and green checkmark. That is the place where you type little equations in order to search for results.

The following are types of things to put in a filter and then run Statistics:

AMT > 100000 Amount is larger than $100,000

BRNCH = "04" All items where branch is 04 or Bakersfield
INVDT < `971231` All items where the invoice date is less than 12/31/97

Other innovative uses:

1. Try running the Statistics Command on a date column. The highest five and lowest five will give you a first indication
of cutoff issues, as well as the oldest item. The following is an excerpt of the output:

Highest: 01/01/99 01/01/99 12/31/98 12/31/98 12/31/98

Lowest: 02/01/94 01/23/95 03/06/95 06/14/95 08/05/96

Note: Highest, Lowest and Average are the only relevant items when running Statistics on a date column. The totals do not
mean anything.

2. Try running the Statistics Command on an interest rate column (i.e., in a loan file). The number of Zeros could be
interesting or possibly the Average. You could filter on branch, agent, etc. to isolate problem areas. Also, the Highest and
Lowest will get you the five highest and lowest rates. With the filter, you could run Statistics on different loan types,
reporting codes and loan-to-value ratios.

3. Try running Statistics on Unit Cost in an inventory. While the totals will not make sense, Highest, Lowest, Zeros and
Average will be interesting.

Page 3
STRATIFY COMMAND

Description of the Stratify Command:

STRATIFY is a command which allows the user to set up ranges (or buckets) and then stratify a population, like A/R
Balance, into those ranges.

For example, if you have a file with A/R invoices and want to know what percentage of A/R invoices are between $0 and
$10,000, $10,001 and $20,000, $20,001 and $30,000, stratify accomplishes this task. Not only does it show counts of
invoices falling into the ranges, it accumulates (totals) a column you specify as well, like A/R amount.

The following is an example of Stratify's output:

amt COUNT <-- % % --> amt

0.00 -> 9,999.99 55769 97.59% 57.21% 48919866.16

10,000.00 -> 19,999.99 889 1.56% 14.30% 12229398.47
20,000.00 -> 29,999.99 259 0.45% 7.26% 6210639.16
30,000.00 -> 39,999.99 75 0.13% 2.99% 2554187.65
40,000.00 -> 49,999.99 44 0.08% 2.25% 1920252.87
50,000.00 -> 59,999.99 25 0.04% 1.59% 1357067.60
60,000.00 -> 69,999.99 16 0.03% 1.21% 1035701.37
70,000.00 -> 79,999.99 14 0.02% 1.19% 1019193.23
80,000.00 -> 89,999.99 12 0.02% 1.17% 998831.11
90,000.00 -> 100,000.00 9 0.02% 1.00% 855022.46
> 100,000.00 33 0.06% 9.84% 8410945.33

57145 100.00%100.00% 85511105.41

57145 of 60917 met the test: amt > 0

Under the column "amt" at the left shows the column which was stratified on. These are the buckets or ranges. Under the
column "amt" on the right, shows the column which was accumulated. To read the output, you could say there were 259
invoices with amounts between $20,000 and $29,999. These 259 invoices are 0.45% of the total number of invoices and
add up to $6,210,639.16 which is 7.26% of the total A/R balance.

To use the Stratify command in ACL:

Click on Analyze and then on Stratify.

A dialog box will show up. Under "Stratify on", point to the field which you want to use to make the buckets. This could
be A/R balance, Loan Balance, Inventory Extension, Interest rate, etc. If there is only one field listed, click on it to select it.
It may appear to be selected but it is not until you click on it.

Under "Accumulate fields", click on the field(s) you want added or "accumulated" see description above.

Under "Minimum" and "Maximum", select your low and high amount. Use round numbers. In the example output above,
0 was selected for the minimum and $100,000 for the maximum. Because ACL is set for 10 intervals, 10 buckets are made
between the low and the high.

Tip: if the Statistics command was previously run on the field you are stratifying on, the low and high generated by
Statistics will be inserted into the Stratify low and high. You should override these values with round numbers otherwise,
the ranges will be meaningless.

Under "If", it is best to segregate positive amounts from the negative amounts, because it makes more sense for the person
reading the output. Assuming the field being stratifying on is called AMT, put AMT > 0 under "IF".

Page 4