This guide provides a walk-through of how to use the official Algorithmia Python Client to call algorithms and manage data through the Algorithmia platform.
Here you will learn how to install the Algorithmia Python Client, work with the Data API by uploading and downloading files, create and update directories and permission types and last, you’ll learn how to call an algorithm that summarizes text files.
To follow along you can create a new Python script or if you’d rather, you can follow the examples in the Python interpreter.
Getting Started with Algorithmia
To get started, first install the Algorithmia Python Client with pip:
Now import the Algorithmia library and create the Algorithmia client:
Now you’re ready to start working with Algorithmia in Python.
Enterprise Users Only: Specifying an On-Premises Endpoint
If you are running the Algorithmia platform on-premises with Algorithmia Enterprise, you can specify the API endpoint when you create the client object:
Alternately, you can ensure that each of your servers interacting with your Algorithmia Enterprise instance have an environment variable named
ALGORITHMIA_API and the client will use it. The fallback API endpoint is always the hosted Algorithmia marketplace service at https://api.algorithmia.com/
Working with Data Using the Data API
This guide will show you how to work with the Hosted Data option on the Algorithmia platform which is available to both algorithm and application developers.
If you wish to follow along working through the example yourself, create a text file that contains any unstructured text such as a chapter from a public domain book or article. We used a chapter from Burning Daylight, by Jack London which you can copy and paste into a text file. Or copy and paste it from here: Chapter One Burning Daylight, by Jack London. This will be used throughout the guide.
Create a Data Collection
This section will show how to create a data collection which is essentially a folder of data files hosted on Algorithmia for free.
First create a data collection called nlp_directory:
A Data URI uniquely identifies files and directories and contains a protocol “data://” and path “YOUR_USERNAME/data_collection”. For more information on the Data URI see the Data API Specification.
Instead of your username you can also use ‘.my’ when calling algorithms. For more information about the ‘.my’ pseudonym check out the Hosted Data Guide.
Work with Directory Permissions
When we created the data collection in the previous code snippet, the default setting is
ReadAcl.my_algos which is a permission type that allows other users on the platform to interact with your data through the algorithms you create if you decide to contribute to algorithm development. This means users can call your algorithm to perform an operation on your data stored in this collection, otherwise the algorithm you created would only work for you.
To begin working with data directory permissions first add these imports:
Next check for the data collection’s permission type and update those permissions to private:
Notice that we changed our data collection to private, which means that only you will be able to read and write to your data collection.
Note that read access that is set to the default
DataMyAlgorithms allows any algorithm you call to have access to your data collection so most often, this is the setting you want when you are calling an algorithm and are an application developer.
For more information on collection-based Access Control Lists (ACLs) and other data collection permissions go to the Hosted Data Guide.
Upload Data to your Data Collection
So far you’ve created your data collection and checked and updated directory permissions. Now you’re ready to upload the text file that you created at the beginning of the guide to your data collection using the Data API.
First create a variable that holds the path to your data collection and the text file you will be uploading:
Next upload your local file to the data collection using the
This endpoint will replace a file if it already exists. If you wish to avoid replacing a file, check if the file exists before using this endpoint.
You can confirm that the file was created by navigating to Algorithmia’s Hosted Data Source and finding your data collection and file.
Downloading Data from a Data Collection
Next check if the file that you just uploaded to data collections exists and then download the contents of that file as a string:
This will get your file as a string, saving it to the variable
Now you’ve seen how to upload a local data file, check if a file exists in a data collection, and download the file contents as a string.
For more methods on how to get a file using the Data API from a data collection go to the API Specification.
Call an Algorithm
Finally we are ready to call an algorithm. In this guide we’ll use the natural language processing algorithm called Summarizer. This algorithm results in a string that is the summary of the text content you pass in as the algorithm’s input.
A single algorithm may have different input and output types, or accept multiple types of input, so consult the algorithm’s description for usage examples specific to that algorithm.
This example shows the summary of the text file which we downloaded from our data collection and set as the variable called
input in the previous code sample.
Create the algorithm object and pass in the variable
This guide used the the first chapter of Jack London’s Burning Daylight and the Summarizer algorithm outputs:
“It was a quiet night in the Shovel. The miners were in from Moseyed Creek and the other diggings to the west, the summer washing had been good, and the men’s pouches were heavy with dust and nuggets. MacDonald grinned and nodded, and opened his mouth to speak, when the front door swung wide and a man appeared in the light.”
If you are interested in learning more about working with unstructured text data check out our guide Introduction to Natural Language Processing.
Your account can make up to 80 Algorithmia requests at the same time (this limit can be raised if needed).
This guide covered installing Algorithmia via pip, uploading and downloading data to and from a user created data collection, checking if a file exists using the Data API, calling an algorithm, and handling errors.
For convenience, here is the whole script available to run: