Random Selection (SAP Library - Inventory Sampling (MM-IM-PI))

Random Selection

Use

In the random selection process, the system draws the elements to count for each stratum.

Please note that the system performs the random selection on the basis of the transactions previously performed. Therefore, the stock management units, stock population, and stratification can no longer be changed.

However, when performing the random selection, the system does not automatically create the physical inventory documents. Therefore, at this point, you still have the option of deciding whether or not to carry out the physical inventory.

The random selection is carried out with the aid of random numbers, which are generated in the system using an internal random number generator. The random number generator uses the linear congruence method and meets the latest research standards in this area.

Activities

Generating the Random Selection

Enter the number of the inventory sampling and the year by choosing Inventory Sampling ® Change. To go to the inventory sampling header screen, choose Continue. Choose Transaction ® Random selection.

Since all previous transactions cannot be changed after the random selection has been generated, the system prompts you to specify whether you want to carry out the random selection. If you select Yes, the system will carry out the following actions:

It selects the stock management units to be counted

It creates a session for the physical inventory documents (also refer to the section Physical Inventory)

It saves the random selection

The system displays a list of the stock management units to be counted and issues the message Random selection executed - Inventory sampling saved.

The header of the list contains the name of the session created.

Inspecting the Random Selection

After you have created a session, you should check whether the random selection is representative of all the layers. To do this, update the inventory sampling (refer to Update). If the system finds that a subsequent random selection is required for the sampling area, although the book inventory balances and values have not changed, it could be due to the following:

In some strata, the random selection does not correspond to the statistical standard split.

For the parameters, the upper value limit you have chosen for the sampling area is too high.

In the lowest stratum, there are many materials with a stock value of zero although, in the random selection for this stratum, a disproportionately high number of materials with a value greater than zero were chosen.

As the system has not yet created any physical inventory documents for the current inventory sampling, you can decide at this point whether you want to carry out a subsequent random selection or cancel and delete this inventory sampling before you have processed the session.

If the subsequent random selection is very high due to the criteria above, it is advisable to cancel and delete the current inventory sampling. You cannot delete an inventory sampling that you have created, just the elements in the stock population. When a user deletes an inventory sampling, he or she enters a comment in the header data and in the stock management units. This is stored in the system.

When creating a new inventory sampling, you can prevent a subsequent random selection by choosing an upper value limit that corresponds to the criteria for optimal stratification (see Strata).

Carrying out a Plausibility Check

With the plausibility check, you can run checks before the inventory sampling count. These give you an idea of how much work will be required and how successful the inventory sampling will be. During the check, the system equals the fictitious count values for the sampling elements with the current book values. It therefore assumes that the count would not have resulted in any inventory differences.

The plausibility check gives you an idea of the quality of the random selection that has been carried out and thereby of the probable success of the inventory sampling. At the same time, the system provides you with an approximate overview of the expected subsequent random selections. If the results of the plausibility check are not satisfactory, you can carry out a new random selection.