elasticsearch/docs/reference/transform/overview.asciidoc

[role="xpack"]
[[transform-overview]]
= {transform-cap} overview
++++
<titleabbrev>Overview</titleabbrev>
++++

You can choose either of the following methods to transform your data:
<<pivot-transform-overview,pivot>> or <<latest-transform-overview,latest>>.

IMPORTANT: All {transforms} leave your source index intact. They create a new
index that is dedicated to the transformed data.

{transforms-cap} are persistent tasks; they are stored in cluster state which 
makes them resilient for node failures. Refer to <<transform-checkpoints>> and 
<<ml-transform-checkpoint-errors>> to learn more about the machinery behind 
{transforms}.


[[pivot-transform-overview]]
== Pivot {transforms}

You can use {transforms} to _pivot_ your data into a new entity-centric index.
By transforming and summarizing your data, it becomes possible to visualize and
analyze it in alternative and interesting ways.

A lot of {es} indices are organized as a stream of events: each event is an
individual document, for example a single item purchase. {transforms-cap} enable
you to summarize this data, bringing it into an organized, more
analysis-friendly format. For example, you can summarize all the purchases of a
single customer.

{transforms-cap} enable you to define a pivot, which is a set of
features that transform the index into a different, more digestible format.
Pivoting results in a summary of your data in a new index.

To define a pivot, first you select one or more fields that you will use to
group your data. You can select categorical fields (terms) and numerical fields
for grouping. If you use numerical fields, the field values are bucketed using
an interval that you specify.

The second step is deciding how you want to aggregate the grouped data. When
using aggregations, you practically ask questions about the index. There are
different types of aggregations, each with its own purpose and output. To learn
more about the supported aggregations and group-by fields, see
<<put-transform>>.

As an optional step, you can also add a query to further limit the scope of the
aggregation.

The {transform} performs a composite aggregation that paginates through all the
data defined by the source index query. The output of the aggregation is stored
in a _destination index_. Each time the {transform} queries the source index, it
creates a _checkpoint_. You can decide whether you want the {transform} to run
once or continuously. A _batch {transform}_ is a single operation that has a
single checkpoint. _{ctransforms-cap}_ continually increment and process
checkpoints as new source data is ingested.

Imagine that you run a webshop that sells clothes. Every order creates a
document that contains a unique order ID, the name and the category of the
ordered product, its price, the ordered quantity, the exact date of the order,
and some customer information (name, gender, location, etc). Your data set
contains all the transactions from last year.

If you want to check the sales in the different categories in your last fiscal
year, define a {transform} that groups the data by the product categories
(women's shoes, men's clothing, etc.) and the order date. Use the last year as
the interval for the order date. Then add a sum aggregation on the ordered
quantity. The result is an entity-centric index that shows the number of sold
items in every product category in the last year.

[role="screenshot"]
image::images/pivot-preview.png["Example of a pivot {transform} preview in {kib}"]

[[latest-transform-overview]]
== Latest {transforms}

beta::[]

You can use the `latest` type of {transform} to copy the most recent documents
into a new index. You must identify one or more fields as the unique key for
grouping your data, as well as a date field that sorts the data chronologically.
For example, you can use this type of {transform} to keep track of the latest
purchase for each customer or the latest event for each host.

[role="screenshot"]
image::images/latest-preview.png["Example of a latest {transform} preview in {kib}"]

As in the case of a pivot, a latest {transform} can run once or continuously. It
performs a composite aggregation on the data in the source index and stores the
output in the destination index. If the {transform} runs continuously, new unique
key values are automatically added to the destination index and the most recent
documents for existing key values are automatically updated at each checkpoint.

[[transform-performance]]
== Performance considerations

{transforms-cap} perform search aggregations on the source indices then index
the results into the destination index. Therefore, a {transform} never takes
less time or uses less resources than the aggregation and indexing processes.

If your {transform} must process a lot of historic data, it has high resource
usage initially--particularly during the first checkpoint.

For better performance, make sure that your search aggregations and queries are
optimized and that your {transform} is processing only necessary data. Consider
whether you can apply a source query to the {transform} to reduce the scope of
data it processes. Also consider whether the cluster has sufficient resources in
place to support both the composite aggregation search and the indexing of its
results.

If you prefer to spread out the impact on your cluster (at the cost of a slower
{transform}), you can throttle the rate at which it performs search and index
requests. Set the `docs_per_second` limit when you <<put-transform,create>> or
<<update-transform,update>> your {transform}. If you want to calculate the
current rate, use the following information from the
{ref}/get-transform-stats.html[get {transform} stats API]:
```
documents_processed / search_time_in_ms * 1000
```