To-mixpanel NPM

to Mixpanel

wat.

toMixpanel is an ETL script in Node.js that provides one-time data migrations from common product analytics tools... to mixpanel.

It implements Mixpanel's /import , $merge and /engage endpoints

It uses service accounts for authentication, and can batch import millions of events and user profiles quickly.

This script is meant to be run locally and requires a JSON file for configuration.

tldr;

git clone https://github.com/ak--47/toMixpanel.git

cd toMixpanel/

npm install

node index.js ./path-To-JSON-config

alternatively:

npx to-mixpanel ./path-To-JSON-config

Detailed Instructions

Install Dependencies

This script uses npm to manage dependencies, similar to a web application.

After cloning the repo, cd into the /toMixpanel and run:

npm install

this only needs to be done once.

Config File

toMixpanel requires credentials for your source and your destination

Here's an example of a configuration file for amplitude => mixpanel:

{
  "source": {
    "name": "amplitude",
    "params": {
      "api_key": "{{amplitude api key}}",
      "api_secret": "{{ amplitude api secret }}",
      "start_date": "2021-09-17",
      "end_date": "2021-09-17"
    },
    "options": {
      "save_local_copy": true,
      "is EU?": false
    }
  },
  "destination": {
    "name": "mixpanel",
    "project_id": "{{ project id }}",
    "token": "{{ project token }}",
    "service_account_user": "{{ mp service account }}",
    "service_account_pass": "{{ mp service secret }}",
    "options": {
      "is EU?": false,
	  "recordsPerBatch": 2000
    }
  }
}

you can find more configuration examples in the repo.

supported sources

amplitude

required params: api_key, api_secret, start_date, end_date, is EU?

mixpanel

that's right! you can use toMixpanel to migrate one mixpanel project to another!

required params: token, secret, start_date, end_date, is EU?, do_events, do_people

options: where (see docs), event (see docs), recordsPerBatch (in destination)

required params: filePath, event_name_col, distinct_id_col, time_col, insert_id_col (note: filePath can be EITHER a path to a CSV file or a folder which contains multiple CSV files)

ga360*

required params: project_id, bucket_name, private_key_id, private_key, client_email, client_id, auth_uri, token_uri, auth_provider_x590_cert_url, client_x509_cert_url options: path_to_data (for large datasets, does line-by-line iteration)

*note: google analytics does not have public /export APIs, so you'll need to export your data to bigQuery first, and then export your bigQuery tables to google cloud storage as JSON. You can then create a service account in google cloud storage which can access the bucket; the above-mentioned values are given to you when you create a service account