> For the complete documentation index, see [llms.txt](https://docs.videc.de/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.videc.de/acron-9.3/en/acron_der_anlagenchronist/datenverdichtung.md).
# Data compression
Data compression takes place in several stages. The compressed data can be repeatedly generated from the available origin data but, depending on the volume of data, that may take hours. The process variables are compressed according to the custom [compression method](#verdichtungsmethoden) you assign.
The data are always re-compressed in full by the ACRON `DBEngine.exe` program whenever you make any changes to the plant configuration or process variable configuration.
## Compression methods
Compression methods indicate the way in which the data are compressed into a process variable, such as when forming the monthly values from the day data.
The following compression methods are available:
#### - Arithmetic median
The arithmetic average is formed from the single values.
#### - Arithmetic median (5-95%)
Calculates the arithmetic average value in the range of 5% to 95% of the measurement values. The arithmetic average is formed from the single values. The lowest and highest 5% of the values are ignored.
#### - Logarithmic median
The logarithmic median compresses average values of logarithmic measurement variables such as pH values. The Base 10 logarithm is used. For process variables with this compression the measuring range should be limited to -100 to +100, as this already corresponds to a regular measuring range of 1E-99 to 1E99.
#### - Weighted average value
In contrast to the arithmetic median, this method takes into account the time for which a variable had a specific value. That is useful whenever values are not recorded cyclically but only sporadically, as in the case of [Delta Event recording](/acron-9.3/en/acron_der_anlagenchronist/datenaufzeichnungsverfahren.md#deltaeventaufzeichnung).
#### - Last value / First value
Only the last or first value of all values is determined for the time frame. Use this compression method where the process variable is a count value and when you want to output the value in ACRON as a meter reading (as opposed to a meter difference).
#### - Floating maximum / Floating minimum
A floating maximum or minimum of all values is formed with the specified time interval. A floating maximum/minimum is not tied to fixed intervals.
Usually an interval of one hour (3600 seconds) is specified. In this case the hour is searched for at which the averaged or totaled single values produce the maximum. This might be the time range from 12:34:10 to 13:34:10 hours for example.
#### - Frequency
In this compression mode the measurement values are searched according to the frequency of their occurrence, and the measurement value which occurred most frequently is carried over into the next stage. This method is used, for example, when analyzing weather charts.
#### - Sum
All single values are totalized.
#### - Median
The median is the value which exactly 50% of measurement values are below and 50% above. It may differ widely from the average.
The median is formed from the following series of numbers: *1 2 3 8 9*
The result is *3*, because two values are less than it and two greater (the average value produces 4.6).
#### - Percentile value calculation
Percentile values are the values where x-percent of all values fall below or (100-x)% of the values fall above. For example, the 20% percentile value is the value below which 20% of all values fall and 80% fall above. The percentages for the variables perc1 and perc2 can be defined globally in plant management. By default, 15% and 85% are predefined. Another alternative to calculate percentile values is to choose the percentile value compression method. An individual percentage can be selected here for the respective process variable; the percentile value is the corresponding val value. If this compression method is selected, for example, in the compression level daily values, the variable dval is the percentile value of the ival values of the corresponding day.
For the percentile value calculation - depending on the result of the product n\*p - the following formulas are used:
n\*p as integer:
xp=½(xnp+ xnp+1)
n\*p as floating-point:
xp= ⌈xnp ⌉
with n = number of values, p = percent as decimal value
⌈xnp ⌉ → The number in the Gaussian bracket is always rounded up.
5 7 8 9 11 12 14 15 17 18 (values sorted ascending)
Percentile 25%
10\* 0.25 = 2.5 → ⌈2.5⌉ rounded up to 3 → xp= x3 → 8
Percentile 80%
10\*0.8 = 8 → xp=½(x₈+ x₉) → ½(15+17) → 16
#### - Standard deviation – totality / sample
The standard deviation measures the spread of a distribution of values. A distinction is made between:
| | |
| -------------------------------- | -------------------------------- |
| *totality* | *sample* |
|  |  |
#### - Four highest values / Four lowest values
This compression method determines the four highest and four lowest values from all input values and makes them available for output in reports in the Processvariable.yval, Processvariable.yval1, Processvariable.yval2 and Processvariable.yval3 fields with the associated times Processvariable.yvaltm, Processvariable.yval1tm, Processvariable.yval2tm and Processvariable.yval3tm.
#### - Daily maximum with min. 14 days interval / Daily minimum with min. 14 days interval
In this compression method the highest/lowest daily value for each month at an interval of at least 14 days from the previous value is determined.
#### - Last value / First value (auto variable/manual variable with output format as text)
Only the last or first value of all values is determined for the time frame. Use this compression method where the process variable is a count value and when you want to output the value in ACRON as a meter reading (as opposed to a meter difference).
#### - Catenating (auto variable/manual variable with output format as text)
The character string of the analysis level (e.g. daily value) is formed by linking the character strings of the lower analysis level (here interval values).
#### - Catenating without duplicates (auto variable/manual variable with output format as text)
This method skips character strings that already exist.
10:00-11:00 A
11:00-12:00 A
12:00-13:00 D
13:00-14:00 C
14:00-15:00 A
15:00-16:00 D
16:00-17:00 B
17:00-18:00 B
18:00-19:00 C
19:00-20:00 B
Catenate:
`AADCADBBCB`
Catenating without duplicates:
`ADCB`
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.videc.de/acron-9.3/en/acron_der_anlagenchronist/datenverdichtung.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.