API to BigQuery: Load Data into Bigquery using Python

Many businesses today use a variety of cloud-based applications for day-to-day business, like Salesforce, HubSpot, Mailchimp, Zendesk, etc. Companies are also very keen to combine this data with other sources to measure key metrics that help them grow. Given most of the cloud applications are owned and run by third-party vendors – the applications expose their APIs to help companies extract the data into a data warehouse – say, Google BigQuery. This blog details out the process you would need to follow to move data from API to BigQuery.

Besides learning about the data migration process from rest API to BigQuery, we’ll also learn about their shortcomings and the workarounds. Let’s get started.

Note: When you connect API to BigQuery, consider factors like data format, update frequency, and API rate limits to design a stable integration.

Method 1: Loading Data from API to BigQuery using Hevo Data

Hevo is the only real-time ELT No-code Data Pipeline platform that cost-effectively automates data pipelines that are flexible to your needs. With integration with 150+ Data Sources (40+ free sources), we help you not only export data from sources & load data to the destinations but also transform & enrich your data, & make it analysis-ready.

Learn more about Hevo

Here are the steps to move data from API to BigQuery using Hevo:

Step 1: Configure REST API as your source

• Click PIPELINES in the Navigation Bar.
• Click + CREATE in the Pipeline List View.
• In the Select Source Type page, select REST API.
• In the Configure your REST API Source page:

• Specify a unique Pipeline Name, not exceeding 255 characters.
• Set up your REST API Source.
• Specify the data root, or the path, from where you want Hevo to replicate the data.
• Select the pagination method to read through the API response. Default selection: No Pagination.

Step 2: Configure BigQuery as your Destination

• Click DESTINATIONS in the Navigation Bar.
• Click + CREATE in the Destinations List View.
• In Add Destination page select Google BigQuery as the Destination type.
• In the Configure your Google BigQuery Warehouse page, specify the following details:

Yes, that is all. Hevo will do all the heavy lifting to ensure that your analysis-ready data is moved to BigQuery, in a secure, efficient, and reliable manner.

To know in detail about configuring REST API as your source, refer to Hevo Documentation.

SIGN UP HERE FOR A 14-DAY FREE TRIAL

Method 2: API to BigQuery ETL Using Custom Code

The BigQuery Data Transfer Service provides a way to schedule and manage transfers from REST API datasource to Bigquery for supported applications. One advantage of the REST API to Google BigQuery is the ability to perform actions (like inserting data or creating tables) that might not be directly supported by the web-based BigQuery interface. The steps involved in migrating data from API to BigQuery are as follows:

1. Getting your data out of your application using API
2. Preparing the data that was extracted from the Application 
3. Loading data into Google BigQuery

Step 1: Getting data out of your application using API

Below are the steps to extract data from the application using API.

Get the API URL from where you need to extract the data. In this article, you will learn how to use Python to extract data from ExchangeRatesAPI.io which is a free service for current and historical foreign exchange rates published by the European Central Bank. The same method should broadly work for any API that you would want to use.

API URL = https://api.exchangeratesapi.io/latest?symbols=USD,GBP. If you click on the URL you will get below result:

{      "rates":{         "USD":1.1215,       "GBP":0.9034    },    "base":"EUR",    "date":"2019-07-17" }

Reading and Parsing API response in Python:

a. To handle API response will need two important libraries

import requests import json 

b. Connect to the URL and get the response

url = "https://api.exchangeratesapi.io/latest?symbols=USD,GBP" response = requests.get(url) 

c. Convert string to JSON format

parsed = json.loads(data)

d. Extract data and print

date = parsed["date"] gbp_rate = parsed["rates"]["GBP"] usd_rate = parsed["rates"]["USD"]

Here is the complete code:

import requests

import json

url = "https://api.exchangeratesapi.io/latest?symbols=USD,GBP"

response = requests.get(url)

data = response.text

parsed = json.loads(data)

date = parsed["date"]

gbp_rate = parsed["rates"]["GBP"]

usd_rate = parsed["rates"]["USD"]

print("On " + date + " EUR equals " + str(gbp_rate) + " GBP")

print("On " + date + " EUR equals " + str(usd_rate) + " USD")

Step 2: Preparing data received from API

There are two ways to load data to BigQuery. 

1. You can save the received JSON formated data on JSON file and then load into BigQuery.
2. You can parse the JSON object, convert JSON to dictionary object and then load into BigQuery.

Step 3: Loading data into Google BigQuery

We can load data into BigQuery directly using API call or can create CSV file and then load into BigQuery table.

• Create a Python script to extract data from API URL and load (UPSERT mode)  into BigQuery table. Here UPSERT is nothing but Update and Insert operations. This means – if the target table has matching keys then update data, else insert a new record.

import requests 
import json from google.cloud
import bigquery url = "https://api.exchangeratesapi.io/latest?symbols=USD,GBP" 
response = requests.get(url) 
data = response.text 
parsed = json.loads(data) 
base = parsed["base"] 
date = parsed["date"] 
client = bigquery.Client() 
dataset_id = 'my_dataset' 
table_id = 'currency_details' 
table_ref = client.dataset(dataset_id).table(table_id) 
table = client.get_table(table_ref) for key, value in parsed.items():        if type(value) is dict:            for currency, rate in value.items():                QUERY = ('SELECT target_currency FROM my_dataset.currency_details where currency=%', currency)                query_job = client.query(QUERY)                               if query_job == 0:                    QUERY = ('update my_dataset.currency_details set  rate = % where currency=%',rate, currency)                    query_job = client.query(QUERY)                else:                    rows_to_insert = [                    (base, currency, 1, rate)                    ]                    errors = client.insert_rows(table, rows_to_insert) assert errors == []

• Load JSON file to BigQuery. You need to save the received data in JSON file and load JSON file to BigQuery table.

import requests import json from google.cloud import bigquery url = "https://api.exchangeratesapi.io/latest?symbols=USD,GBP" response = requests.get(url) data = response.text parsed = json.loads(data) for key, value in parsed.items():        if type(value) is dict:            with open('F:Pythondata.json', 'w') as f:                json.dump(value, f) client = bigquery.Client(project="analytics-and-presentation") filename = 'F:Pythondata.json' dataset_id = ‘my_dayaset’' table_id = 'currency_rate_details' dataset_ref = client.dataset(dataset_id) table_ref = dataset_ref.table(table_id) job_config = bigquery.LoadJobConfig() job_config.source_format = bigquery.SourceFormat.NEWLINE_DELIMITED_JSON job_config.autodetect = True with open(filename, "rb") as source_file:    job = client.load_table_from_file(source_file, table_ref, job_config=job_config) job.result()  # Waits for table load to complete. print("Loaded {} rows into {}:{}.".format(job.output_rows, dataset_id, table_id)) 

Limitations of writing custom scripts and developing ETL to load data from API to BigQuery

1. The above code is written based on the current source as well as target destination schema. If the data coming in is either from the source or the schema on BigQuery changes, ETL process will break.
2. In case you need to clean your data from API – say transform time zones, hide personally identifiable information and so on, the current method does not support it. You will need to build another set of processes to accommodate that. Clearly, this would also need you to invest extra effort and money.
3. You are at a serious risk of data loss if at any point your system breaks. This could be anything from source/destination not being reachable to script breaks and more. You would need to invest upfront in building systems and processes that capture all the fail points and consistently move your data to the destination.
4. Since Python is an interpreted language, it might cause performance issue to extract from API and load data into BigQuery api. 
5. For many APIs, we would need to supply credentials to access API. It is a very poor practice to pass credentials as a plain text in Python script. You will need to take additional steps to ensure your pipeline is secure. 

API to BigQuery: Use Cases

• Advanced Analytics: BigQuery has powerful data processing capabilities that enable you to perform complex queries and data analysis on your API data. This way, you can extract insights that would not be possible within API alone.
• Data Consolidation: If you’re using multiple sources along with API, syncing them to BigQuery can help you centralize your data. This provides a holistic view of your operations, and you can set up a change data capture process to avoid discrepancies in your data.
• Historical Data Analysis: API has limits on historical data. However, syncing your data to BigQuery allows you to retain and analyze historical trends.
• Scalability: BigQuery can handle large volumes of data without affecting its performance. Therefore, it’s an ideal solution for growing businesses with expanding API data.
• Data Science and Machine Learning: You can apply machine learning models to your data for predictive analytics, customer segmentation, and more by having API data in BigQuery.
• Reporting and Visualization: While API provides reporting tools, data visualization tools like Tableau, PowerBI, and Looker (Google Data Studio) can connect to BigQuery, providing more advanced business intelligence options. If you need to convert an API table to a BigQuery table, Airbyte can do that automatically.

Conclusion

From this blog, you will understand the process you need to follow to load data from API to BigQuery. This blog also highlights various methods and their shortcomings. Using these two methods you can move data from API to BigQuery. However, using Hevo, you can save a lot of your time!

You can also check out this blog to learn about connecting Hubspot to BigQuery.

Visit our Website to Explore Hevo

Hevo offers a faster way to move data from databases or SaaS applications into your data warehouse to be visualized in a BI tool. Hevo is fully automated and hence does not require you to code. You can also have a look at the unbeatable Hevo Pricing that will help you choose the right plan for your business needs!

Want to take Hevo for a spin? Sign Up for a 14-day free trial and experience the feature-rich Hevo suite first hand.