pynsee gives a quick access to more than 150 000 macroeconomic series, a dozen datasets of local data, numerous sources available on insee.fr, geographical limits of administrative areas taken from IGN as well as key metadata and SIRENE database containing data on all French companies. Have a look at the detailed API page portail-api.insee.fr.

This package is a contribution to reproducible research and public data transparency. It benefits from the developments made by teams working on APIs at INSEE and IGN.

Installation & API subscription

Credentials are necessary to access SIRENE API available through pynsee by the module sirene. API credentials can be created here : portail-api.insee.fr. All other modules are freely accessible.

```python

Download Pypi package

pip install pynsee[full]

Get the development version from GitHub

git clone https://github.com/InseeFrLab/pynsee.git

cd pynsee

pip install .[full]

Subscribe to portail-api.insee.fr and get your credentials!

Save your credentials with init_conn function :

from pynsee.utils import initconn initconn(sirenekey="mysirene_key")

Beware : any change to the keys should be tested after having cleared the cache

Please do : from pynsee.utils import clearallcache; clearallcache()

```

Data Search and Collection Advice

• Macroeconomic data : First, use get_dataset_list to search what are your datasets of interest and then get the series list with get_series_list. Alternatively, you can make a keyword-based search with search_macrodata, e.g. search_macrodata('GDP'). Then, get the data with get_dataset or get_series
• Local data : use first get_local_metadata, then get data with get_local_data
• Metadata : e.g. function to get the classification of economic activities (Naf/Nace Rev2) get_activity_list
• Sirene (French companies database) : use first get_dimension_list, then use search_sirene with dimensions as filtering variables
• Geodata : get the list of available geographical data with get_geodata_list and then retrieve it with get_geodata
• Files on insee.fr: get the list of available files on insee.fr with get_file_list and then download it with download_file

For further advice, have a look at the documentation and gallery of the examples.

Example - Population Map

```python import math

import matplotlib.cm as cm import matplotlib.pyplot as plt import numpy as np import pandas as pd

from pynsee.geodata import getgeodatalist, get_geodata

get geographical data list

geodatalist = getgeodata_list()

get departments geographical limits

mapcom = getgeodata("ADMINEXPRESS-COG-CARTO.LATEST:commune").tocrs(epsg=3035)

area calculations depend on crs which fits metropolitan france but not overseas departements

figures should not be considered as official statistics

mapcom.attrs["area"] = mapcom.area / 10**6 mapcom = mapcom.to_crs(epsg=3857)

mapcom['REFAREA'] = 'D' + mapcom['codeinsee'] mapcom['density'] = mapcom['population'] / mapcom.attrs["area"]

mapcom = mapcom.transform_overseas(departement=['971', '972', '974', '973', '976', 'NR'], factor=[1.5, 1.5, 1.5, 0.35, 1.5, 1.5])

mapcom = mapcom.zoom( departement=["75","92", "93", "91", "77", "78", "95", "94"], factor=1.5, startAngle = math.pi * (1 - 3.5 * 1/9))

density_ranges = [ 40, 80, 100, 120, 150, 200, 250, 400, 600, 1000, 2000, 5000, 10000, 20000 ]

rvals = np.full(len(mapcom), "< 40", dtype=object)

list_ranges = ["< 40"]

for rmin, rmax in zip(densityranges, densityranges[1:]): rangestring = f"[{rmin}, {rmax}[" listranges.append(range_string)

rvals[(mapcom.density >= rmin) & (mapcom.density < rmax)] = range_string

rvals[mapcom.density.values > density_ranges[-1]] = "> 20 000"

list_ranges.append("> 20 000")

mapcom.loc[:, "range"] = pd.Categorical(rvals, ordered=True, categories=list_ranges)

fig, ax = plt.subplots(1, 1, figsize=(15, 15)) lgd = {'bboxtoanchor': (1.1, 0.8), 'title': 'density per km2'} mapcom.plot(column="range", cmap=cm.viridis, legend=True, ax=ax, legendkwds=lgd) ax.setaxis_off() ax.set(title='Distribution of population in France') plt.show()

fig.savefig('popfrance.svg', format='svg', dpi=1200, bboxinches = 'tight', pad_inches = 0) ```

How to avoid proxy issues ?

```python

Use the proxyserver argument of the initconn function to change the proxy server address

from pynsee.utils import initconn initconn(sirenekey="mysirenekey", httpproxy="http://myproxyserver:port", httpsproxy="http://myproxy_server:port")

Beware : any change to the keys should be tested after having cleared the cache

Please do : from pynsee.utils import *; clearallcache()

Alternativety you can use directly environment variables as follows.

Beware not to commit your credentials!

import os os.environ['sirenekey'] = 'mysirenekey' os.environ['httpproxy'] = "http://myproxyserver:port" os.environ['httpsproxy'] = "http://myproxy_server:port"

```

Support

Feel free to open an issue with any question about this package using the Github repository.

Contributing

All contributions, whatever their forms, are welcome. See CONTRIBUTING.md