# Data migration

Memgraph Lab provides multiple data migration options, including importing
predefined datasets for exploration and tools for importing custom data. Users
can import CSV files using the [CSV file import tool](#csv-file-import), import
their dataset with Cypher queries using the
[`CYPHERL`](https://memgraph.com/docs/memgraph-lab/getting-started/data-migration#cypherl) format or
[export](#export-to-cypherl) their data to `CYPHERL` file. They can also [stream
real-time data](#streams) from Kafka, Redpanda and Pulsar. The following
sections provide details on how to use each approach efficiently.

## Datasets

In Memgraph Lab, there is a dedicated **Datasets** section that offers a
collection of datasets covering different topics and sizes and allows you to
explore and experiment with a variety of preloaded datasets. It's designed to
help you:
- **Explore the Cypher query language**: Run queries on real-world data and
  practice Cypher.
- **Learn Memgraph Lab features**: Understand visualization, query execution and
  database exploration tools.
- **Test and experiment**: Experiment with different graph structures before
  working with your own data. 

You can check the structure of the dataset by checking its graph schema, as well
as reading the explanations of all the entities and their properties.

![Datasets](https://memgraph.com/docs/pages/data-visualization/lab-user-manual/dataset.png)

### Exploring datasets

Once you got Memgraph Lab [up and
running](https://memgraph.com/docs/memgraph-lab/getting-started#installation-and-deployment), select a
dataset to work with in the **Datasets** section of the sidebar. After importing
a dataset, you can:
- [Download query collection](https://memgraph.com/docs/memgraph-lab/features/collections) - Download a set of pre-prepared Cypher queries
  to familiarize with the dataset. 
- [Run Cypher queries](https://memgraph.com/docs/memgraph-lab/querying) – Execute queries in the **Query Execution** window to
  analyze your data using Cypher. 
- [View the graph schema](https://memgraph.com/docs/memgraph-lab/features/graph-schema) – Understand the structure and relationships within
  the dataset.
- [Explore datasets](https://memgraph.com/docs/querying/exploring-datasets) with graph analytics – Follow premade
  tutorials and interact with the dataset.

By leveraging the Datasets section in Memgraph Lab, users can quickly gain
hands-on experience with graph databases, making it easier to transition into
solving real-world graph problems with Memgraph.

## Data migration tools

Memgraph Lab provides multiple tools for importing data into the database, each
suited for different use cases. The [CSV file import](#csv-file-import) tool allows
users to load data from CSV files, making it easier to structure nodes and
relationships. For real-time data ingestion, Memgraph supports
[streaming](#streams) Kafka, Redpanda and Pulsar sources. Additionally, the
[CYPHERL](#cypherl) format enables importing data using Cypher queries,
allowing direct creation of nodes and relationships. Each method offers specific
advantages depending on the dataset size, structure and import requirements.

### CSV file import

Memgraph Lab provides the [CSV file import
tool](https://memgraph.com/docs/memgraph-lab/features/csv-file-import) that allows users to load
data row by row from CSV files, simplifying the process of adding nodes and
relationships to the graph database. Users can merge the new data with the
existing dataset or reset the database entirely before importing. While CSV file
import tool is more suitable for small datasets and quick imports, larger
datasets should be handled using the `LOAD CSV` command directly. Check out
[import best practices](https://memgraph.com/docs/data-migration/best-practices) for more details on how
to do that.

![](https://memgraph.com/docs/pages/data-migration/csv/csv_import.png)

#### Import CSV files

- [Importing files with and without
headers](https://memgraph.com/docs/memgraph-lab/features/csv-file-import#import-files-with-and-no-header): Memgraph supports CSV files with or without headers. Users can upload
files, configure nodes by assigning labels and properties and manage indexing
and uniqueness constraints. Relationships can be configured by selecting the
start and end nodes and setting relevant properties.
- [Importing data from a single CSV
file](https://memgraph.com/docs/memgraph-lab/features/csv-file-import#import-data-from-a-single-csv-file): Data can be imported from a single CSV file containing both nodes and
relationships. Users assign labels, set properties and configure relationships
accordingly.
- [Importing data from multiple CSV
Files](https://memgraph.com/docs/memgraph-lab/features/csv-file-import#import-data-from-multiple-csv-files): Data can be distributed across multiple CSV files, with each file
containing nodes of a specific label or relationships of a specific type. Users
configure each file by assigning labels, selecting properties and specifying
handling rules for duplicate data.

#### Finalize the import

Once all files are configured, Memgraph executes the import process in four
steps:
1. **Validating files** to ensure consistency.
2. **Uploading to Memgraph** for processing.
3. **Executing the import**, including setting up indexes and constraints.
4. **Cleaning up unnecessary files** after a successful import.

After completing the import, users can visualize the data using Cypher queries,
such as:

```cypher
MATCH p=()-[]-() RETURN p;
```

### CYPHERL

Memgraph Lab provides a convenient way to [import](#import-from-cypherl) or
[export](#export-to-cypherl) your data using the `CYPHERL` format, which
represents data in the form of Cypher queries.

#### Import from CYPHERL

To import a `CYPHERL` file in Memgraph Lab you need to navigate to the **Import** section, select the `CYPHERL` file or drag
and drop it into Memgraph Lab. `CYPHERL` files contain Cypher queries (such as
`CREATE` and `MERGE` statements) that create nodes and relationships in your
graph database. The benefit of using `CYPHERL` is that you only need a single
file to import both nodes and relationships.

When importing `CYPHERL` files via Memgraph Lab, be
aware of the following: 
- You can import up to 1 million nodes and relationships via Memgraph Lab using
`CYPHERL` files.
- All existing indexes are automatically dropped during the import process To
ensure optimal performance, include `CREATE INDEX` queries at the beginning of
your `CYPHERL` file.
- To speed up import time, consider creating indexes on appropriate nodes or
node properties.

In a `CYPHERL` file, each Cypher query must be written in a new row. Here's an
example of what a `CYPHERL` file might contain:

```cypher
CREATE (:Person {id: "100", name: "Daniel", age: 30, city: "London"}); 
CREATE (:Person {id: "101", name: "Alex", age: 15, city: "Paris"}); 
MATCH (u:Person), (v:Person) 
WHERE u.id = "100" AND v.id = "101" 
CREATE (u)-[:IS_FRIENDS_WITH]->(v);
```

This example creates two `Person` nodes and establishes a relationship between
them.

#### Export to CYPHERL

In addition to importing, Memgraph Lab also allows you to export your data in
`CYPHERL` format. This is useful for backing up your database or transferring
data between Memgraph instances Memgraph Lab user manual.

![](https://memgraph.com/docs/pages/data-visualization/lab-user-manual/cypherl-export.png)

### Streams

[Streaming in Memgraph Lab](https://memgraph.com/docs/memgraph-lab/features/streams) enables real-time data processing, allowing users to
ingest and analyze streaming data. With built-in tools and integrations, you can
manage and monitor streams within Memgraph Lab, with constant data flow and
insights. Here are some of the features streams offer:

- **Data ingestion**: Connect to streaming platforms like Kafka, Redpanda and Pulsar or create custom streams
that suit your specific needs. Live data from various sources flows directly
into your applications, enabling real-time processing with minimal delays to
keep your system responsive and up to date. 
- **Stream processing**: Run Cypher queries on incoming data to extract valuable insights
instantly. Triggers can be set up to react the moment new data arrives, enabling
automated, event-driven workflows that respond in real time. As data streams in,
it can be modified and enriched on the fly, ensuring it’s properly structured
and ready for use throughout your system. 
- **Stream management**: Monitor all active streams in real time. You can start,
stop, or adjust streams as needed to adapt to changing demands, while also
fine-tuning performance to keep data flowing efficiently and avoid processing
bottlenecks.

To explore how to set up and manage streams in Memgraph Lab, check out the full
documentation for more detailed explanations: [Managing streams in Memgraph
Lab](https://memgraph.com/docs/memgraph-lab/features/streams).

> **Info**
>
> To make sure you achieve the best import speed with Memgraph, please refer to the [import best
> practices](https://memgraph.com/docs/data-migration/best-practices).