cepan

Retrieves data from aws cost explore as a pandas dataframe.

Main features

Support for input with type hints
Retrieving results as pandas.Dataframe

Installation

pip install cepan

Usage

from datetime import datetime

import cepan as ce

df = ce.get_cost_and_usage(
    time_period=ce.TimePeriod(
        start=datetime(2020, 1, 1),
        end=datetime(2020, 1, 2),
    ),
    granularity="DAILY",
    filter=ce.And(
        [
            ce.Dimensions(
                "SERVICE",
                ["Amazon Simple Storage Service", "AmazonCloudWatch"],    
            ),
            ce.Tags("Stack", ["Production"]),
        ]
    ),
    metrics=["BLENDED_COST"],
    group_by=ce.GroupBy(
        dimensions=["SERVICE", "USAGE_TYPE"],
    ),
)
print(df)

All paginated results will be returned as a Dataframe.

          Time                        SERVICE  BlendedCost
0   2020-01-01  Amazon Simple Storage Service   100.000000
1   2020-01-01               AmazonCloudWatch    10.000000

List of currently supported APIs

get_dimension_values
get_tags
get_cost_and_usage

Alias of aws service name

Normally, the Cost Explorer API requires complex and long names to filter by service name. For example, if you only need the value of an EC2 instance, you would need to specify Amazon Elastic Compute Cloud - Compute.

df = ce.get_cost_and_usage(
    time_period=ce.TimePeriod(
        start=datetime(2020, 1, 1),
        end=datetime(2020, 1, 2),
    ),
    granularity="DAILY",
    filter=ce.Dimensions(
        "SERVICE",
        ["Amazon Elastic Compute Cloud - Compute"],    
    ),
    group_by=ce.GroupBy(
        dimensions=["SERVICE", "USAGE_TYPE"],
    ),
)

cepan supports aliases with short service names. If you only need the value of the EC2 instance, you can specify it with EC2.

df = ce.get_cost_and_usage(
    time_period=ce.TimePeriod(
        start=datetime(2020, 1, 1),
        end=datetime(2020, 1, 2),
    ),
    granularity="DAILY",
    filter=ce.Dimensions(
        "SERVICE",
        ["EC2"],    
    ),
    group_by=ce.GroupBy(
        dimensions=["SERVICE", "USAGE_TYPE"],
    ),
)

Correspondence table of aliases is shown in service_alias.tsv. You can also run the show_service_alias method to get the table.

print(ce.show_service_alias())

License