canalyst-candas

Canalyst Data Science Library

Canalyst's Python SDK is intended as a full featured data science library facilitating dataframe manipulation, data model mapping, and interactive scenario analysis of core Model, ModelSet, and ModelMap objects.

Python Version Requirement

This package requires Python version >=3.7. Recommended Python version >=3.8.

Usage

Install Required Packages

Run pip install -r requirements.txt to install the requried packages.

To Preview Sample Data

The Python SDK supports a preview of sample data for three tickers: DIS US, TSLA US, or NFLX US.

import canalyst_candas as cd
model = cd.SampleModel(ticker="DIS US") 
df = model.model_frame()

Configuration

The Python SDK supports three ways of providing configuration, in order of precedence:

1. Creating a Config instance
2. Using environment variables
3. Using a configuration file

Config is used to store a user's configuration information. On initial canalyst_candas import, it will attempt to retrieve the configuration.

Using Config

A Config instance can be created with the desired configurations.

In Python/iPython or Jupyter Notebook. Replace <..> with values

import sys
sys.path.append('<path to sdk-python>/src')
import canalyst_candas as cd
from canalyst_candas.configuration.config import Config

# if you want to set a default directory, otherwise `default_dir` will be a temporary directory
# from pathlib import Path
# home = Path.home()
# default_dir = f"{home}/<any place under your home directory>"

config_info = {
  "canalyst_api_key": "<Canalyst API key>",
  "canalyst_s3_id": "<S3 Access Key ID>",
  "canalyst_s3_key": "<S3 Secret Key>",
  "fred_key": "<Fred key>",
  "default_dir": "",
  "mds_host": "",
  "wp_host": "",
  "verify_ssl": "True",
}

config = Config(config=config_info)

ticker = "<ticker you want test with>"
model = cd.Model(ticker=ticker, config=config)
model_set = cd.ModelSet(ticker_list=[ticker], config=config)
cd.ModelMap(
  ticker=ticker,
  config=config,
  time_series_name=time_series,
  col_for_labels=<label>,
  common_size_tree=True,
  notebook=False,
  auto_download=False,
)

Using Environment Variables

Environment variables can be set in your terminal. The SDK looks for

• CANALYST_API_KEY
• S3_ACCESS_KEY_ID
• S3_SECRET_KEY
• FRED_KEY (optional)
• DEFAULT_DIR (optional; default is a temporary directory)
• MDS_HOST (optional; default is production Model Data System)
• WP_HOST (optional; default is production Web Portal)
• VERIFY_SSL (optional: default is "True")

In Python/iPython or Jupyter Notebook. Replace <..> with values

import sys
sys.path.append('<path to sdk-python>/src')
import canalyst_candas as cd

ticker = "<ticker to test with>"
time_series = "<time series to test with>"
label = "<label of interest>
model = cd.Model(ticker=ticker)
model_set = cd.ModelSet(ticker_list=[ticker])
cd.ModelMap(
  ticker=ticker,
  time_series_name=time_series,
  col_for_labels=<label>,
  common_size_tree=True,
  notebook=False,
  auto_download=False,
)

Using a Configuration File

When canalyst_candas is imported and if there is not already an existing configuration file, a configuration file, keys.json, is created in the user's home directory, C:\Users\<username>\canalyst\keys.json on Windows or /User/<username>/canalyst/keys.json on a Unix-based OS (e.g. Mac, Linux). The configuration file can be moved but must be under a user's home directory and must be in directly under a 'canalyst' folder. Examples of valid locations:

Windows

• C:\Users\<username>\canalyst\keys.json
• C:\Users\<username\Downloads\canalyst\keys.json'

Unix-based OS

• /Users/<username/canalyst/keys.json
• /Users/<username>/Downloads/canalyst/keys.json

The contents of keys.json:

{
    "canalyst_api_key": "",
    "canalyst_s3_id": "",
    "canalyst_s3_key": "",
    "fred_key": "",
    "default_dir": "",
    "mds_host": "",
    "wp_host": "",
    "verify_ssl": "",
    "proxies": ""
}

Fill in

1. canalyst_api_key: Canalyst API token. Get it at https://app.canalyst.com/settings/api-tokens
2. canalyst_s3_id: AWS S3 Access Key ID
3. canalyst_s3_key: AWS S3 Secret Key
4. (Optional) default_dir: your chosen directory for where downloads will go. The default is a temporary directory.

In Python/iPython or Jupyter Notebook. Replace <..> with values

import sys
sys.path.append('<path to sdk-python>/src')
import canalyst_candas as cd

ticker = "<ticker to test with>"
time_series = "<time series to test with>"
label = "<label of interest>
model = cd.Model(ticker=ticker)
model_set = cd.ModelSet(ticker_list=[ticker])
cd.ModelMap(
  ticker=ticker,
  time_series_name=time_series,
  col_for_labels=<label>,
  common_size_tree=True,
  notebook=False,
  auto_download=False,
)

Proxy Support

The Candas library allows for proxies to be passed in as part of the request. The proxies are to be formatted as a python dictionary:

{"http": "http://http_proxy.com:8000", "https": "https://https_proxy.com:8001"}

If you only have one proxy, the http and https keys can have the same proxy link. This dictionary is to be passed in as part of the config object:

    "proxies": {
        "http": "http://18.206.117.131:3128",
        "https": "http://18.206.117.131:3128"
      }