Skip to content

![earthmover](assets/ea-earthmover.png)

earthmover transforms collections of tabular source data (flat files, FTP files, database tables/queries) into text-based (JSONL, XML) data via YAML configuration.

![earthmover demo](assets/earthmover_demo.gif)

Quick-start

  1. Install earthmover with 

    pip install earthmover

  2. Create an earthmover.yml configuration file that defines your project config, data sources, transformations, and destinations:

    earthmover.yml
     1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
    version: 2

config:
  output_dir: ./output/

sources:
  csv_source:
    file: ./data/file.csv
    header_rows: 1
  sql_source:
    connection: "postgresql://user:pass@host/database"
    query: >
      select column1, column2, column3
      from myschema.mytable

transformations:
  stacked:
    source: $sources.csv_source
    operations:
      - operation: union
        sources:
          - $sources.sql_source

destinations:
  mydata:
    source: $transformations.stacked
    template: ./json_templates/mydata.jsont
    extension: jsonl
    linearize: True

  3. Create the ./json_templates/mydata.jsont template file (which may use Jinja) to use when rendering the data for your mydata destination:

    ./json_templates/mydata.jsont
    {
    "column1": "{{column1}}",
    "column2": "{{column2}}",
    "column3": "{{column3}}"
}

  4. Now run earthmover 

    earthmover run
    and look for the output file output/mydata.json.

How it works

earthmover is similar to dbt, though it executes data transformations locally using dataframes (rather than in a SQL engine). Like dbt, it creates a data dependency DAG from your earthmover.yml configuration and materializes output data in dependency-order.

Read more

Above is a simple quick-start example. Please read the documentation for more information about earthmover's many features, including: