Extraction of International Financial Statistics data from the IMF

The example below retrieves quarterly (period: Q) Seasonally Adjusted Real Gross Domestic Product (indicator: NGDP_R_SA_XDC) for the USA (country code: US), from the International Financial Statistics (IFS) series using the IMFDataPy package. The function call returns the observation values, and the time period for each value (in the format YYYY-MM-DD).

First, we begin with loading the IMFDataPy library for data extraction and from the IMF and pandas for data manipulation.

`IMFDataPy` package

Source code foris available on Github.

The package can be installed using pip. !pip install imfdatapy

from imfdatapy.imf import *

import pandas as pd # for QoQ change calculation

To download the data from the International Financial Statistics, we use the IFS class, provding the search terms for the index we are looking for, the country code, the period frequency (Q) and the period. Use the download_data method to download the data and the metadata to ‘../out’ folder and create a pandas dataframe. The log messages specify which files are created in the ‘../out’ directory.

pd.options.display.max_colwidth = 90

ifs = IFS(search_terms=["gross domestic product"], countries=["US"], period='Q', start_date="2010",
                  end_date="2023")
df = ifs.download_data()
df

2022-12-01 00:10:26,551 imf.py:117 - INFO - Output all IMF series table to ../out/series_imf.csv
2022-12-01 00:10:26,569 imf.py:119 - INFO - Output series containing 'IFS' table to ../out/series_ifs.csv
2022-12-01 00:10:39,190 imf.py:249 - WARNING - Failed to download CL_FREQ.
2022-12-01 00:10:39,770 imf.py:132 - INFO - Output dimension CL_AREA_IFS table to ../out/dim_cl_area_ifs.csv
2022-12-01 00:10:39,784 imf.py:132 - INFO - Output dimension CL_INDICATOR_IFS table to ../out/dim_cl_indicator_ifs.csv
2022-12-01 00:10:40,153 imf.py:149 - INFO - Output meta data of IFS table to ../out/meta_ifs.csv
2022-12-01 00:10:40,172 imf.py:151 - INFO - Output meta data of IFS containing 'gross domestic product' table to ../out/meta_gross domestic product_US_Q_2010_2023.csv
2022-12-01 00:10:40,482 imf.py:171 - INFO - Output data of IFS containing 'gross domestic product' table to ../out/data_gross domestic product_US_Q_2010_2023.csv

	Description	Country	Period	Value	ID
0	Gross Domestic Product, Deflator, Seasonally Adjusted, Index	US	2010-01-01	99.3392105021347	NGDP_D_SA_IX
1	Gross Domestic Product, Deflator, Seasonally Adjusted, Index	US	2010-04-01	99.8236914454449	NGDP_D_SA_IX
2	Gross Domestic Product, Deflator, Seasonally Adjusted, Index	US	2010-07-01	100.125182516395	NGDP_D_SA_IX
3	Gross Domestic Product, Deflator, Seasonally Adjusted, Index	US	2010-10-01	100.711915536026	NGDP_D_SA_IX
4	Gross Domestic Product, Deflator, Seasonally Adjusted, Index	US	2011-01-01	101.231913593634	NGDP_D_SA_IX
...	...	...	...	...	...
97	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2021-07-01	5887605	NGDP_SA_XDC
98	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2021-10-01	6087280.3	NGDP_SA_XDC
99	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-01-01	6185120	NGDP_SA_XDC
100	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-04-01	6312119	NGDP_SA_XDC
101	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-07-01	6415822.3	NGDP_SA_XDC

255 rows × 5 columns

  <script>
    const buttonEl =
      document.querySelector('#df-beed8f22-19d1-48ce-94f1-44112fd49b9f button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-beed8f22-19d1-48ce-94f1-44112fd49b9f');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

Here, all the data that matched the search term is loaded. To view the index names, use the meta data file as shown below.

meta = pd.read_csv('../out/meta_gross domestic product_US_Q_2010_2023.csv')

meta

	ID	Description.@xml:lang	Description
0	NGDP_D_IX	en	Gross Domestic Product, Deflator, Index
1	NGDP_D_SA_IX	en	Gross Domestic Product, Deflator, Seasonally Adjusted, Index
2	NGDP_XDC	en	Gross Domestic Product, Nominal, Domestic Currency
3	NGDP_SA_XDC	en	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency
4	NGDP_NSA_XDC	en	Gross Domestic Product, Nominal, Unadjusted, Domestic Currency
5	NGDP_R_XDC	en	Gross Domestic Product, Real, Domestic Currency
6	NGDP_R_SA_XDC	en	Gross Domestic Product, Real, Seasonally Adjusted, Domestic Currency
7	NGDP_R_NSA_XDC	en	Gross Domestic Product, Real, Unadjusted, Domestic Currency

  <script>
    const buttonEl =
      document.querySelector('#df-6e4d3215-86e4-42cf-ae20-dc99625e5931 button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-6e4d3215-86e4-42cf-ae20-dc99625e5931');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

We are interested in Gross Domestic Product, Real, Seasonally Adjusted, Domestic Currency. We will filter the dataframe to contain only this index.

df = df[df['ID']=='NGDP_SA_XDC'].reset_index()

df.tail(n=5)

	index	Description	Country	Period	Value	ID
46	97	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2021-07-01	5887605	NGDP_SA_XDC
47	98	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2021-10-01	6087280.3	NGDP_SA_XDC
48	99	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-01-01	6185120	NGDP_SA_XDC
49	100	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-04-01	6312119	NGDP_SA_XDC
50	101	Gross Domestic Product, Nominal, Seasonally Adjusted, Domestic Currency	US	2022-07-01	6415822.3	NGDP_SA_XDC

  <script>
    const buttonEl =
      document.querySelector('#df-8591373e-e28e-4bab-bf8e-8ccb702676ff button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-8591373e-e28e-4bab-bf8e-8ccb702676ff');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

df['Value'] = pd.to_numeric(df['Value'])

df['QoQ'] = df['Value'].pct_change()

Now, we may plot the results using matplotlib.

import matplotlib.pyplot as plt

plt.rcParams.update({'font.size': 15})

t = df['Period']
data1 = df['Value'] * 10**6
data2 = df['QoQ']

labels = [f'Q{int(ts.month/3)+1}\n{ts.year}' if ts.month == 1
          else f'Q{int(ts.month/3)+1}' for ts in t]

fig, ax1 = plt.subplots()

ax1.set_xlabel('Period')
ax1.set_ylabel('Real GDP', color='blue')
ax1.set_xticks(t)
ax1.set_xticklabels(labels);
ax1.plot(t, data1, color='blue')
ax1.tick_params(axis='y', labelcolor='black')

ax2 = ax1.twinx()

ax2.set_ylabel('QoQ change', color='red')
ax2.set_xticks(t)
ax2.set_xticklabels(labels);
ax2.plot(t, data2, '--', color='red')
ax2.tick_params(axis='y', labelcolor='black')

fig.set_size_inches(25.5, 5.5)
plt.title('Real GDP by Quarter in the US')
fig.tight_layout()
plt.show()

png

Extraction of International Financial Statistics data from the IMF

**IMFDataPy** package

`IMFDataPy` package