Skip to main content

Package to calculate economic complexity and associated variables

Project description

Economic Complexity and Product Complexity

by the Growth Lab at Harvard's Center for International Development

This package is part of Harvard Growth Lab’s portfolio of software packages, digital products and interactive data visualizations. To browse our entire portfolio, please visit To learn more about our research, please visit Harvard Growth Lab’s home page.


Python package to calculate economic complexity indices.

STATA implementation of the economic complexity index available at:

Explore complexity and associated data using Harvard CID's Atlas tool:


Installation: At terminal: pip install ecomplexity


from ecomplexity import ecomplexity
from ecomplexity import proximity

# Import trade data from CID Atlas
data_url = ""
data = pd.read_csv(data_url, compression="zip", low_memory=False)
data = data[['year','location_code','hs_product_code','export_value']]

# Calculate complexity
trade_cols = {'time':'year', 'loc':'location_code', 'prod':'hs_product_code', 'val':'export_value'}
cdata = ecomplexity(data, trade_cols)

# Calculate proximity matrix
prox_df = proximity(data, trade_cols)


data: pandas dataframe containing production / trade data.
    Including variables indicating time, location, product and value
cols_input: dict of column names for time, location, product and value.
    Example: {'time':'year', 'loc':'origin', 'prod':'hs92', 'val':'export_val'}
presence_test: str for test used for presence of industry in location.
    One of "rca" (default), "rpop", "both", or "manual".
    Determines which values are used for M_cp calculations.
    If "manual", M_cp is taken as given from the "value" column in data
val_errors_flag: {'coerce','ignore','raise'}. Passed to pd.to_numeric
    *default* coerce.
rca_mcp_threshold: numeric indicating RCA threshold beyond which mcp is 1.
    *default* 1.
rpop_mcp_threshold: numeric indicating RPOP threshold beyond which mcp is 1.
    *default* 1. Only used if presence_test is not "rca".
pop: pandas df, with time, location and corresponding population, in that order.
    Not required if presence_test is "rca" (default).
continuous: Used to calculate product proximities, indicates whether
    to consider correlation of every product pair (True) or product
    co-occurrence (False). *default* False.
asymmetric: Used to calculate product proximities, indicates whether
    to generate asymmetric proximity matrix (True) or symmetric (False).
    *default* False.


  • There are very minor differences in the values of density, COI and COG between STATA and Python due to the way matrix computations are handled by the two. These should be aligned in the future.
  • knn options for density: in the future, allow knn parameter for density calculation


Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ecomplexity-0.5.2.tar.gz (9.1 kB view hashes)

Uploaded source

Built Distribution

ecomplexity-0.5.2-py2-none-any.whl (11.3 kB view hashes)

Uploaded py2

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page