Employment by sector and departement

from pynsee.localdata import get_local_metadata, get_nivgeo_list, get_local_data
from pynsee.geodata import get_geodata
from pynsee.metadata import get_activity_list
from shapely import wkt

import pandas as pd
from pandas.api.types import CategoricalDtype
import geopandas as gpd
import seaborn as sns
import math

from matplotlib.pyplot import pie
import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1.inset_locator import inset_axes

# get a list all data available : datasets and variables
metadata = get_local_metadata()

# geographic metadata
nivgeo = get_nivgeo_list()

# get departement map
dep = get_geodata('ADMINEXPRESS-COG-CARTO.LATEST:departement')

dep = dep.translate(departement=['971', '972', '974', '973', '976'],
                   factor=[1.5, 1.5, 1.5, 0.3, 1.5])

dep = dep.zoom(factor = 3.5 , startAngle = math.pi * (1 - 3 * 1/9))

dep = dep[["insee_dep", "geometry"]]
dep = dep.rename(columns={'insee_dep': 'CODEGEO'})

# get local data from INSEE API
data = get_local_data(dataset_version = 'GEO2020FLORES2017',
                     variables =  'NA17',
                     nivgeo = 'DEP')
data = data.rename(columns={'NA17': 'A17'})

Existing environment variables used, instead of locally saved credentials
Getting data: 100%|██████████| 101/101 [01:21<00:00,  1.23it/s]

# keep only headcounts and remove the total from the data
df = data[data["UNIT"] == "NBSALET"].reset_index(drop=True)
df = df[~df["A17"].isin(["ENS"])].reset_index(drop=True)

# compute share of activity in total employment by department
df['employment_share'] = 100 * df["OBS_VALUE"] / df['OBS_VALUE'].groupby(df['CODEGEO']).transform('sum')

# merge values and geographic limits
datamap = df.merge(dep, how = 'left', on = 'CODEGEO')
datamap = gpd.GeoDataFrame(datamap)

# add english activity labels at A17 level
activity = get_activity_list("A17")
activity["TITLE_A17_EN"] = activity["A17"] + " - " + activity["TITLE_A17_EN"]
datamap = datamap.merge(activity, how='left', on='A17')

# compute center location for each departement
def get_center(x):
    return x.centroid.wkt

datamap["center"] = datamap['geometry'].apply(lambda x: get_center(x))

# list of departements
list_dep = datamap["CODEGEO"].unique()

# make color palette
colors = sns.color_palette("Set1") + sns.color_palette("Set2")

# order activity sectors according to average share
emplTot = datamap[["employment_share", "TITLE_A17_EN"]].groupby(['TITLE_A17_EN'], observed=True).mean()
emplTot = emplTot.sort_values(by=["employment_share"],  ascending=False)

datamap["TITLE_A17_EN"] = datamap["TITLE_A17_EN"].astype(CategoricalDtype(categories=emplTot.index, ordered=True))
datamap = datamap.sort_values(["TITLE_A17_EN"]).reset_index(drop=True)
datamap

	CODEGEO	NIVGEO	UNIT	UNIT_label_fr	A17	OBS_VALUE	NA17_label	DATASET_VERSION	DATASET_NAME	DATA_DATE	GEO_DATE	CODEGEO_label	employment_share	geometry	A5	A10	TITLE_A17_EN	TITLE_A17_FR	center
0	24	DEP	NBSALET	Postes salariés dans les établissements	OQ	47091	Administration publique, enseignement, santé h...	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Dordogne	39.785237	MULTIPOLYGON (((4538.510 5624971.758, 4634.387...	OQ	OQ	OQ - Public administration and defence, educat...	Administration publique, enseignement, santé h...	POINT (82503.88583124451 5638329.410666983)
1	33	DEP	NBSALET	Postes salariés dans les établissements	OQ	199158	Administration publique, enseignement, santé h...	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Gironde	32.943017	MULTIPOLYGON (((-128762.061 5717199.240, -1287...	OQ	OQ	OQ - Public administration and defence, educat...	Administration publique, enseignement, santé h...	POINT (-64206.77633210372 5594297.759162161)
2	71	DEP	NBSALET	Postes salariés dans les établissements	OQ	61315	Administration publique, enseignement, santé h...	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Saône-et-Loire	34.317841	MULTIPOLYGON (((531046.916 5928690.377, 531251...	OQ	OQ	OQ - Public administration and defence, educat...	Administration publique, enseignement, santé h...	POINT (505725.92242259247 5884496.094662105)
3	01	DEP	NBSALET	Postes salariés dans les établissements	OQ	54202	Administration publique, enseignement, santé h...	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Ain	29.047782	MULTIPOLYGON (((533820.757 5815685.538, 533972...	OQ	OQ	OQ - Public administration and defence, educat...	Administration publique, enseignement, santé h...	POINT (595403.9693304703 5796511.958803039)
4	22	DEP	NBSALET	Postes salariés dans les établissements	OQ	66843	Administration publique, enseignement, santé h...	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Côtes-d'Armor	36.247933	MULTIPOLYGON (((-382905.380 6257752.911, -3828...	OQ	OQ	OQ - Public administration and defence, educat...	Administration publique, enseignement, santé h...	POINT (-318942.6429392173 6180631.785546785)
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1763	13	DEP	NBSALET	Postes salariés dans les établissements	C2	1446	Cokéfaction et raffinage	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Bouches-du-Rhône	0.182878	MULTIPOLYGON (((597782.742 5338896.723, 597787...	BE	BE	C2 - Manufacture of coke and refined petroleum...	Cokéfaction et raffinage	POINT (566095.8521214168 5395484.572187278)
1764	26	DEP	NBSALET	Postes salariés dans les établissements	C2	0	Cokéfaction et raffinage	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Drôme	0.0	MULTIPOLYGON (((540386.992 5605296.851, 540431...	BE	BE	C2 - Manufacture of coke and refined petroleum...	Cokéfaction et raffinage	POINT (575197.8809542124 5572442.056683063)
1765	59	DEP	NBSALET	Postes salariés dans les établissements	C2	269	Cokéfaction et raffinage	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Nord	0.027313	MULTIPOLYGON (((338476.314 6474007.637, 339089...	BE	BE	C2 - Manufacture of coke and refined petroleum...	Cokéfaction et raffinage	POINT (357483.2519709124 6524765.489153067)
1766	27	DEP	NBSALET	Postes salariés dans les établissements	C2	0	Cokéfaction et raffinage	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Eure	0.0	MULTIPOLYGON (((35180.302 6327852.729, 34839.6...	BE	BE	C2 - Manufacture of coke and refined petroleum...	Cokéfaction et raffinage	POINT (110910.78782311625 6294369.484342183)
1767	33	DEP	NBSALET	Postes salariés dans les établissements	C2	0	Cokéfaction et raffinage	GEO2020FLORES2017	Fichier localisé des rémunérations et de l'emp...	2017	2020	Gironde	0.0	MULTIPOLYGON (((-128762.061 5717199.240, -1287...	BE	BE	C2 - Manufacture of coke and refined petroleum...	Cokéfaction et raffinage	POINT (-64206.77633210372 5594297.759162161)

1768 rows × 19 columns

emplTot

	employment_share
TITLE_A17_EN
OQ - Public administration and defence, education, human health and social work activities	36.017645
GZ - Wholesale and retail trade; repair of motor vehicles and motorcycles	13.79726
MN - Professional, scientific, technical, administrative and support service activities	9.349464
C5 - Other manufacturing	6.822302
FZ - Construction	6.040776
HZ - Transportation and storage	5.366087
IZ - Accommodation and food service activities	4.07249
RU - Other services activities	3.733203
C1 - Manufacture of food products, beverages and tobacco products	3.016751
KZ - Financial and insurance activities	2.961577
JZ - Information and communication	1.843239
C3 - Manufacture of electrical, computer and electronic equipment; Manufacture of machinery	1.714648
DE - Mining and quarrying; energy, water supply, sewerage, waste management and remediation activities	1.60692
AZ - Agriculture, forestry and fishing	1.422385
C4 - Manufacture of transport equipment	1.343475
LZ - Real estate activities	0.872148
C2 - Manufacture of coke and refined petroleum products	0.01963

# draw plot

def extract_lon(x):
    return wkt.loads(x).xy[0][0]
def extract_lat(x):
    return wkt.loads(x).xy[1][0]

ax = datamap.plot(color='white', edgecolor='black', figsize = (15,15))

ax_sub_coeff = 80000

for idep in list_dep:

    df = datamap[datamap["CODEGEO"] == idep].reset_index(drop=True)
    df["lon"] = df["center"].apply(lambda x: extract_lon(x))
    df["lat"] = df["center"].apply(lambda x: extract_lat(x))

    df = df[df["lon"] == min(df["lon"])].reset_index(drop=True)
    center = df.loc[0,"lon"], df.loc[0,"lat"]

    ratios = df["employment_share"]

    ax_sub = ax.inset_axes(
                    bounds=(center[0]-ax_sub_coeff/2,
                            center[1]-ax_sub_coeff/2,
                            ax_sub_coeff, ax_sub_coeff),
                    transform=ax.transData
                   )

    ax_sub.pie(ratios, wedgeprops={'edgecolor': 'gray','linewidth': 0.25}, colors = colors)


plt.suptitle('Employment by sector and departement', x = 0.4, y = 0.85,
            horizontalalignment= 'left',
            fontsize = 12, fontweight='bold')
plt.legend(emplTot.index, loc="lower left", bbox_to_anchor=(-27, -9), ncol=2)
plt.savefig('empl_map_fr.png')
plt.close()

from PIL import Image
def resize_pix(img, pct):
    hsize = int((float(img.size[1])*float(pct)))
    wsize = int((float(img.size[0])*float(pct)))
    img = img.resize((wsize,hsize), Image.Resampling.LANCZOS)
    return img

img = Image.open('empl_map_fr.png').convert("RGB")
img = resize_pix(img, pct=0.5)
img