How to Load Data from CSV to BigQuery: 4 Easy Methods

Databases support CSV files on a regular basis. BigQuery, Google’s data warehouse as a service, combines data storage and analytics in one package. BigQuery allows you to conduct real-time SQL queries on billions of records. Stakeholders are always looking to find faster and better ways to get all their data from all their sources into BigQuery.

A very popular and semantically easy data format these days is CSV. It can store data from databases, clickstreams, browsing trails, social media interactions, page views, and a multitude of sources. In this article, you will see 4 ways to move data from CSV to BigQuery. Read along to select the method that works best for your business!

What is BigQuery?

Google BigQuery is a fully managed, serverless cloud data warehouse designed for high performance and scalability. It uses Dremel technology, which optimizes SQL queries through a tree structure and stores data in a columnar format for efficient processing. This architecture allows users to analyze massive datasets quickly without managing infrastructure.

BigQuery supports ANSI SQL, making it easy for users to write queries and integrate with BI tools for data visualization. With BigQuery ML, users can train and run machine learning models using SQL. Its ability to handle petabytes of data makes it ideal for analytics and enterprise-level data processing.

Why move data from CSV to BigQuery?

• Enhances overall efficiency: Uploading CSV files to BigQuery simplifies data management and enhances the efficiency of your analytical workflows, making it easier to handle and analyze large datasets.
• Performance: BigQuery is designed to handle massive volumes of data efficiently, offering quick query execution that reduces the time needed to gain insights from your data.
• Advanced Analytics: BigQuery provides advanced analytics tools, such as ML and spatial data analysis, which deliver deeper insights to inform wise decision-making.
• Cost-Effective: BigQuery’s pay-as-you-go pricing approach ensures you only pay for the storage and queries you use, eliminating the need for expensive hardware or software.
• Scalability: Features like real-time analytics, on-demand storage scaling, BigQuery ML, and optimization tools make it easier to manage and scale your data analysis processes as needed.

What Are The Methods To Load Data From CSV to BigQuery?

Method 1: Using Hevo Data

Step 1: Configure Google Sheets as your Source.

Step 2: Configure BigQuery as your Destination details

After the pipeline has completed ingesting, you can preview your loaded table by opening BigQuery and previewing it.

Method 2: Using Command Line Interface

Step 1: Google Cloud Setup

Step 1. a) Install the gcloud command line interface

Step 1. b) Authenticate yourself to Google Cloud. For that, run a command:

gcloud auth login

Step 1. c) Sign in with your account and give all permissions to Google Cloud SDK.

After you complete all these steps, this window should appear:

Step 2: Command Prompt Configurations

Step 2. a) Run command:

bq 

This will enter you in Google’s big query.

Step 2. b) To check the number of datasets in your projects, run the command:

bq ls <project_id>:

Step 2. c) Now, we need to create a new dataset. In order to do so, you can run the command:

Bq mk <new_dataset_name>

Step 3: Load the data into the dataset

Step 3. a) First, go to your cloud SDK directory and add the CSV file that you want to upload. In my case, this is the directory.

Step 3. b) Now, to load the file, go back to your command line and run the following command:

bq load --source_format=CSV example_cli.bank demo.csv ID: integer,NAME:string,AGE:integer

The output of this command will be Upload Complete.

Step 4: Preview of the data

Step 4. a) To check whether the table has been created, you can use the following command:

bq ls example_cli

Step 4. b) If you want to see the schema of the table, you can do so by using the command:

bq show <dataset_name>.<table_name>

Step 4. c) To preview your table, go back to BigQuery, go to tables, and click on preview.

Method 3: Using BigQuery Web

Step 1: Create a new Dataset

To create a new dataset in BigQuery, go to your BigQuery studio, click on the three dots beside your project ID, and click Create Dataset.

Step 2: Create a Table

To create a new table, click on the three dots next to your dataset name and click on Create Table.

Note:

• Keep the file format as CSV.
• Turn on auto-detect. This will automatically detect the incoming table schema and generate a schema accordingly.
• Change create table from to Upload. Upload the file that you want to load into BigQuery.

Step 3: Preview Table 

Step 3. a) You can edit the schema you just created. To do so, click on the table name and click on the schema tab. You can make changes to the schema as you like from here, and finally, click on edit schema.

Step 3. b) To preview the table, click on the preview tab beside the details tab.

Method 4: Using Web API

Step 1: Configure BigQuery API

Go to Google Cloud Console and look for APIs and services. Search for BigQuery and click on BigQuery API.

Note: Make sure that the service is enabled.

Step 2: Configuring the Python script

Open your code editor and type the given Python script.

import time

from google.cloud import bigquery

import os

from google.cloud import bigquery

os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = ""

client = bigquery.Client()

job_config = bigquery.LoadJobConfig(

    source_format=bigquery.SourceFormat.CSV,

    skip_leading_rows=1,

    autodetect=True,

)

table_id = ""

with open(r'demo.csv', "rb") as source_file:

    job = client.load_table_from_file(source_file, table_id, job_config=job_config)

while job.state != 'DONE':

    time.sleep(2)

    job.reload()

    print(job.state)

print(job.result())

table = client.get_table(table_id)

print(

    "Loaded {} rows and {} columns to {}".format(

        table.num_rows, len(table.schema), table_id

    )

)

Note: 

• In Google application credentials, provide your service account’s json key.
• In table_id. Give a new table name for what you want to create.

Step 4: Preview the table

You can now preview the table you created by following the steps mentioned in this blog.

Limitations of Moving Data from CSV to BigQuery

• Nesting and repetitive data are not supported in CSV files.
• BOM (byte order mark) characters should be removed. They may result in unanticipated consequences.
• BigQuery cannot read the data in parallel if you use gzip compression. Importing compressed CSV data into BigQuery takes longer than loading uncompressed data. See Loading compressed and uncompressed data for further information.
• You can’t use the same load job to load compressed and uncompressed files.
• A gzip file can be up to 4 GB in size.

Learn More on How to Unload and Load CSV to Redshift and Import CSV File Into PostgreSQL Table

Conclusion

Loading CSV data into BigQuery is a powerful way to enhance your data management and analytical capabilities. Whether dealing with large datasets, requiring real-time analytics, or leveraging advanced analytics tools like BigQuery ML, importing your CSV data into BigQuery can significantly streamline your workflows. This article provides a step-by-step guide for setting up an import CSV to BigQuery connection using four different methods.

The first three methods are manual and can become time-consuming. Also, writing custom scripts requires high-level code knowledge, which not everyone may be familiar with. To avoid the manual part and automate the entire process, you can always look up Hevo Data and sign up for a 14-day free trial.

FAQ

Share your thoughts on loading data from CSV to BigQuery in the comments!