Supply-chain analyses#

This notebook shows one end-to-end workflow for using MARIO to study a supply chain with EXIOBASE. Starting from yearly parser inputs, the same Database instance is used to move from structural indicators to bilateral trade matrices, embodied trade content, contributor breakdowns, concentration, and exposure.

The example focuses on Nickel ores and concentrates, but the same pattern can be reused for any sector or commodity.

[23]:

import mario

1. Define the EXIOBASE inputs and the focal supply chain#

The workflow below assumes that you have a local folder with yearly EXIOBASE archives. The first year initializes the MARIO database, and the remaining years are imported as additional scenarios so that the same supply chain can be compared through time.

[ ]:

FOLDER = 'YOUR_PATH'
TABLE = "IOT"
YEARS = range(2013, 2015)
SYSTEM = "pxp"
SECTOR = "Nickel ores and concentrates"

# Uncomment if you need to download the dataset first.
# mario.download_exiobase(version="3.10.1", system=SYSTEM, years=YEARS, path=FOLDER)

2. Parse the first year and import the rest as scenarios#

This keeps the whole time series inside one Database. That matters because all the supply-chain indicators below can then be compared with scenario="all" or by selecting one baseline year.

[25]:

db = mario.parse_exiobase(
    path=f"{FOLDER}/{TABLE}_{YEARS[0]}_{SYSTEM}.zip",
    table="IOT",
    unit="Monetary",
)
db.rename_scenario("baseline", YEARS[0])

INFO Parser: reading EXIOBASE IOT from /var/folders/7_/vg3_ld6n2pd73xyzt9dcdhlc0000gn/T/mario_exiobase_iot_aacpo95a.
INFO Parser: reading EXIOBASE IOT extensions from /var/folders/7_/vg3_ld6n2pd73xyzt9dcdhlc0000gn/T/mario_exiobase_iot_aacpo95a.
INFO Parser: using split extension layout with 7 extension directories.
INFO Parser: EXIOBASE IOT parsed with 200 sectors, 9 value-added rows and 725 extension rows.
INFO Metadata: initialized.

[ ]:

for year in YEARS[1:]:
    print(f"Processing year {year}...")
    db.parse_exiobase(
        path=f"{FOLDER}/{TABLE}_{year}_{SYSTEM}.zip",
        table="IOT",
        unit="Monetary",
        new_scenario=year,
    )
    print("DONE")

Processing year 2014...
INFO Parser: reading EXIOBASE IOT from /var/folders/7_/vg3_ld6n2pd73xyzt9dcdhlc0000gn/T/mario_exiobase_iot_176gbwyq.
INFO Parser: reading EXIOBASE IOT extensions from /var/folders/7_/vg3_ld6n2pd73xyzt9dcdhlc0000gn/T/mario_exiobase_iot_176gbwyq.
INFO Parser: using split extension layout with 7 extension directories.
INFO Parser: EXIOBASE IOT parsed with 200 sectors, 9 value-added rows and 725 extension rows.
INFO Parser: state payload ready with 5 canonical blocks.
DONE

3. Start from structural position: `calc_linkages(...)`#

A complete supply-chain analysis should begin with the sector’s structural role in the production network. calc_linkages(...) tells you whether the focal sector is mainly a strong demander of upstream inputs, a strong downstream enabler, or both.

In multi-regional mode, MARIO also reports Local Share and Foreign Share, which help separate linkage intensity from linkage geography. In aggregated mode, it adds amplification ratios that show whether indirect effects are much larger than the direct ones.

[27]:

linkages = db.calc_linkages(scenario="2013")
linkages.head(10)

INFO Resolver: resolving w for 2013.
INFO Resolver: trying w via formula build_iot_w_from_z.
INFO Resolver: resolved w via formula build_iot_w_from_z.
INFO Resolver: resolving b for 2013.
INFO Resolver: trying b via formula build_iot_b_from_X_Z.
INFO Resolver: resolved b via formula build_iot_b_from_X_Z.
INFO Resolver: resolving g for 2013.
INFO Resolver: trying g via formula build_iot_g_from_b.
INFO Resolver: resolved g via formula build_iot_g_from_b.
WARNING Normalization not available for multi-regional mode.

[27]:

			Total Forward		Total Backward		Direct Forward		Direct Backward		Total Forward		Total Backward		Direct Forward		Direct Backward
			Local	Foreign	Local	Foreign	Local	Foreign	Local	Foreign	Local Share	Foreign Share	Local Share	Foreign Share	Local Share	Foreign Share	Local Share	Foreign Share
Region	Level	Item
AT	Sector	Additives/Blending Components	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
		Air transport services (62)	0.354104	0.390661	0.408775	0.817409	0.185977	0.105890	0.238681	0.270432	0.475458	0.524542	0.333372	0.666628	0.637198	0.362802	0.468817	0.531183
		Aluminium and aluminium products	0.696550	1.513491	0.781241	1.069240	0.487980	0.428916	0.406256	0.295364	0.315175	0.684825	0.422183	0.577817	0.532209	0.467791	0.579026	0.420974
		Aluminium ores and concentrates	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
		Animal products nec	0.098513	0.432467	0.668477	0.573460	0.070839	0.189544	0.374608	0.147690	0.185530	0.814470	0.538254	0.461746	0.272056	0.727944	0.717230	0.282770
		Anthracite	0.029730	2.908958	1.001713	0.605283	0.000000	0.932988	0.533770	0.172782	0.010117	0.989883	0.623345	0.376655	0.000000	1.000000	0.755457	0.244543
		Ash for treatment, Re-processing of ash into clinker	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
		Aviation Gasoline	0.536250	1.253700	0.379844	1.128521	0.390954	0.547946	0.237192	0.524693	0.299589	0.700411	0.251825	0.748175	0.416395	0.583605	0.311323	0.688677
		BKB/Peat Briquettes	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
		Basic iron and steel and of ferro-alloys and first products thereof	0.042204	2.440849	0.708880	0.848255	0.011841	0.944595	0.404283	0.245636	0.016997	0.983003	0.455246	0.544754	0.012381	0.987619	0.622051	0.377949

[28]:

aggregated_linkages = db.calc_linkages(
    scenario="2013",
    multi_mode=False,
 )

aggregated_linkages.head(10)

[28]:

			Total Forward	Total Backward	Direct Forward	Direct Backward	Forward Amplification	Backward Amplification
Region	Level	Item
AT	Sector	Additives/Blending Components	0.000000	0.000000	0.000000	0.000000	NaN	NaN
		Air transport services (62)	0.602529	1.219589	0.574110	1.140037	2.551728	2.408470
		Aluminium and aluminium products	1.787964	1.840527	1.803558	1.571110	2.410352	2.637439
		Aluminium ores and concentrates	0.000000	0.000000	0.000000	0.000000	NaN	NaN
		Animal products nec	0.429572	1.235257	0.512181	1.169562	2.039222	2.377831
		Anthracite	2.377452	1.598352	1.835213	1.582155	3.149759	2.274418
		Ash for treatment, Re-processing of ash into clinker	0.000000	0.000000	0.000000	0.000000	NaN	NaN
		Aviation Gasoline	1.448102	1.500252	1.846841	1.706059	1.906433	1.979780
		BKB/Peat Briquettes	0.000000	0.000000	0.000000	0.000000	NaN	NaN
		Basic iron and steel and of ferro-alloys and first products thereof	2.008835	1.548759	1.881335	1.455337	2.596152	2.395891

4. Map the bilateral geography of the chain with `calc_trades(...)`#

Once you know whether the sector is structurally important, the next question is geographic: where do the relevant trade flows actually run? calc_trades(...) answers that by building origin-by-destination trade matrices.

The examples below keep the split between intermediate and final trade, show totals, and then aggregate the origin side to continents. The same method can also build an animated plot with scenario="all".

[29]:

split_trade = db.calc_trades(
    item=SECTOR,
    scenario="2013",
    aggregate=False,
    total=True,
    auto_open=False,
    path=False,
    show_plot=False,
)

split_trade.head(10)

[29]:

	Intermediate										...	Final
Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	TR	TW	US	WA	WE	WF	WL	WM	ZA	Total
Region
AT	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.000000
AU	0.0	2096.333717	0.008412	0.0	42.148222	71.910781	0.0	1378.824431	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.011084	0.0	14.873840
BE	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.000000
BG	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.000000
BR	0.0	0.000000	0.087826	0.0	576.120518	0.118718	0.0	4.715478	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	16.708715
CA	0.0	0.000000	0.000000	0.0	0.000000	1105.895311	0.0	1.166070	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.027729	0.0	0.756920
CH	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.000000
CN	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	2374.228873	0.0	0.000000	...	0.0	0.0	3.310822e-14	0.0	0.0	0.0	0.0	0.000000	0.0	0.211057
CY	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.000000
CZ	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.008093	...	0.0	0.0	0.000000e+00	0.0	0.0	0.0	0.0	0.000000	0.0	0.147689

10 rows × 100 columns

[31]:

origin_clustered_trades = db.calc_trades(
    item=SECTOR,
    scenario="2013",
    clusters="continent",
    clusters_direction="origin",
    total=True,
    auto_open=False,
    path=False,
    show_plot=False,
)
origin_clustered_trades

[31]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	TR	TW	US	WA	WE	WF	WL	WM	ZA	Total
Region
Europe	0.0	0.000000	0.908103	0.0	0.000000	51.909231	0.0	160.407913	9.757916e-13	1.659926	...	0.002653	0.0	0.002786	0.072842	132.738378	15.831240	0.000000	0.000000	4.469047	2236.514003
Asia	0.0	70.382138	0.018823	0.0	0.000000	0.000000	0.0	5673.822022	0.000000e+00	0.000000	...	26.589091	0.0	71.057130	841.185627	52.852502	0.003665	4.146893	236.196678	0.000000	9182.657505
America	0.0	0.000000	0.087826	0.0	592.822909	1106.740844	0.0	5.881573	0.000000e+00	0.000000	...	0.000000	0.0	0.006253	0.000000	56.187196	0.000000	318.642076	2.372726	0.000000	2156.621242
Oceania	0.0	2097.503758	0.008412	0.0	55.834461	71.910781	0.0	1378.830230	0.000000e+00	0.000000	...	0.000000	0.0	0.000000	0.015612	0.000000	0.000000	0.000000	0.948428	0.000000	3639.674247
Africa	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	21.008060	0.000000e+00	0.000000	...	0.000000	0.0	0.000000	0.000000	0.000000	1089.456976	0.000000	0.000000	448.475774	1581.516306
Total	0.0	2167.885896	1.023165	0.0	648.657370	1230.560856	0.0	7239.949798	9.757916e-13	1.659926	...	26.591744	0.0	71.066169	841.274081	241.778076	1105.291881	322.788969	239.517833	452.944821	18796.983303

6 rows × 50 columns

[32]:

all_trades = db.calc_trades(
    item=SECTOR,
    scenario="all",
    show_plot=True,
    auto_open=False,
    path=False,
    title=f"{SECTOR} - all scenarios",
)

If you want a shareable figure, the same method can also export HTML or static images.

db.calc_trades(
    item=SECTOR,
    scenario=base_scenario,
    save_plot=f"{FOLDER}/nickel_trade_{base_label}.html",
    auto_open=False,
 )

In practice this block answers the first geographic question of the analysis: which origin-destination pairs actually carry the focal supply chain?

5. Add embodied indicators with `calc_trades_content(...)`#

Trade values alone are not enough if the question is about value added, emissions, energy, water, labour, or other extensions. calc_trades_content(...) lets you attach those indicators directly to the same trade geometry.

Here indicator means the quantity embodied in the traded flow. It can be one factor-of-production row, one satellite-account row, or the generic "total value added" shortcut.

The method argument changes the analytical meaning of the result:

direct: only the content generated directly by the production node that supplies the traded item
total: direct content plus all indirect upstream rounds required to produce that traded item
upstream: only the indirect upstream contribution, so it isolates supplier-chain dependence net of the direct producer itself

A useful way to read them is: if total is much larger than direct, the traded flow carries an important hidden upstream chain. If upstream is large, the indicator is not mainly generated by the last exporting sector, but by its wider supplier network.

The first example below uses one factor-of-production row and focuses on the upstream part of the chain.

[33]:

factor_indicator = db.get_index("Factor of production")[0]

upstream_value_added = db.calc_trades_content(
    indicator=factor_indicator,
    item=SECTOR,
    scenario="2013",
    method="upstream",
    clusters="continent",
    clusters_direction="origin",
    show_plot=True,
    auto_open=False,
    path=False,
    title=f"{factor_indicator} embodied in {SECTOR} trade",
)
upstream_value_added

INFO Resolver: resolving v for 2013.
INFO Resolver: trying v via formula build_iot_v_from_V_X.
INFO Resolver: resolved v via formula build_iot_v_from_V_X.
INFO Resolver: resolving m for 2013.
INFO Resolver: trying m via formula build_iot_m_from_v_w (compute_method=auto, runtime=inverse).
INFO Resolver: resolved m via formula build_iot_m_from_v_w (compute_method=auto, runtime=inverse).

[33]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region
Europe	0.0	0.000000	0.051146	0.0	0.000000	1.705098	0.0	3.271703	1.128445e-14	0.018337	...	0.000381	0.000031	0.0	0.000053	0.000965	1.535070	0.520025	0.000000	0.000000	0.146799
Asia	0.0	0.470155	0.000102	0.0	0.000000	0.000000	0.0	69.824418	0.000000e+00	0.000000	...	0.000000	0.594932	0.0	0.384311	4.671407	0.657557	0.000020	0.050877	2.917291	0.000000
America	0.0	0.000000	0.006531	0.0	44.081402	19.193387	0.0	0.370852	0.000000e+00	0.000000	...	0.000000	0.000000	0.0	0.000040	0.000000	0.359475	0.000000	2.039046	0.041134	0.000000
Oceania	0.0	21.327568	0.000086	0.0	0.567729	0.731194	0.0	14.020044	0.000000e+00	0.000000	...	0.000000	0.000000	0.0	0.000000	0.000159	0.000000	0.000000	0.000000	0.009644	0.000000
Africa	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.171755	0.000000e+00	0.000000	...	0.000000	0.000000	0.0	0.000000	0.000000	0.000000	8.950772	0.000000	0.000000	5.049428

5 rows × 49 columns

[35]:

all_value_added_content = db.calc_trades_content(
    indicator="total value added",
    item=SECTOR,
    scenario="2013",
    method="total",
    show_plot=True,
    auto_open=False,
    path=False,
    title=f"Total value added embodied in {SECTOR} trade",
)
all_value_added_content.head(10)

[35]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region
AT	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0
AU	0.0	2097.503759	0.008412	0.0	55.834461	71.910781	0.0	1378.830231	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.015612	0.0	0.0	0.000000	0.948428	0.0
BE	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0
BG	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0
BR	0.0	0.000000	0.087826	0.0	592.822909	0.118718	0.0	4.715497	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.005586	0.000000	0.0
CA	0.0	0.000000	0.000000	0.0	0.000000	1106.622126	0.0	1.166075	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.004880	2.372726	0.0
CH	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0
CN	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	2374.439931	0.0	0.000000	...	0.000000	0.0	0.0	0.002177	0.135476	0.0	0.0	2.401789	0.000000	0.0
CY	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0
CZ	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.000000	0.0	0.155746	...	0.012579	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000	0.000000	0.0

10 rows × 49 columns

6. Explain the result with `breakdown=True`, concentration, and exposure#

A total embodied-content matrix tells you how much indicator content is carried by each trade pair, but not where that content originates inside the production system. breakdown=True keeps the contributor detail so you can identify the production nodes behind the result.

In practice, the breakdown table answers: who is actually generating the embodied content behind this observed trade flow? The first row level still identifies the trade origin Region, while the deeper levels identify the contributing production nodes. If you sum those contributor levels back together, you recover the collapsed trade-content matrix for the same method.

The last two indicators are summaries built on top of that contributor table, and they answer different questions:

calc_trades_concentration(...): how concentrated is the contributor base behind each trade pair?
calc_trades_exposure(...): how much of that contributor base depends on one selected Region or group of Regions?

calc_trades_concentration(...) is computed as an HHI on contributor-Region shares. A value closer to 1 means that only a few contributor Regions dominate the embodied content. Lower values mean the same content is spread across a broader set of contributor Regions. This is a diversification indicator, not a size indicator: a small trade flow and a large trade flow can have the same concentration if their contributor shares are distributed in the same way.

calc_trades_exposure(...) instead returns a share between 0 and 1. It measures how much of the embodied content is attributable to the Regions selected in exposed_to. High exposure means strong dependence on those Regions even if the observed exporter-importer map looks diversified. In other words, direct trade partners and hidden upstream dependency are not necessarily the same thing.

[36]:

value_added_breakdown = db.calc_trades_content(
    indicator="total value added",
    item=SECTOR,
    scenario="2013",
    method="upstream",
    breakdown=True,
    clusters="continent",
    clusters_direction="origin",
)

value_added_breakdown.head(10)

[36]:

			Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region	Region	Level	Item
Europe	AT	Sector	Additives/Blending Components	0.0	0.0	0.000000e+00	0.0	0.0	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	...	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	0.000000e+00	0.0	0.0	0.000000e+00
			Air transport services (62)	0.0	0.0	1.370348e-05	0.0	0.0	2.198436e-03	0.0	2.263640e-03	1.906428e-17	1.388419e-05	...	3.152899e-07	5.183499e-08	0.0	2.440290e-08	3.793416e-07	2.593339e-03	6.704903e-04	0.0	0.0	1.892747e-04
			Aluminium and aluminium products	0.0	0.0	1.291321e-05	0.0	0.0	5.328156e-04	0.0	4.721795e-03	2.558788e-17	6.381986e-05	...	2.394068e-06	6.957241e-08	0.0	1.012454e-07	9.711813e-07	3.480750e-03	1.624877e-04	0.0	0.0	4.586913e-05
			Aluminium ores and concentrates	0.0	0.0	0.000000e+00	0.0	0.0	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	...	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	0.000000e+00	0.0	0.0	0.000000e+00
			Animal products nec	0.0	0.0	4.824001e-08	0.0	0.0	4.659533e-06	0.0	1.266859e-05	1.662091e-19	4.165766e-07	...	1.423709e-08	4.519158e-10	0.0	2.036641e-10	5.926079e-09	2.260960e-05	1.421067e-06	0.0	0.0	4.011573e-07
			Anthracite	0.0	0.0	8.834852e-10	0.0	0.0	5.914017e-09	0.0	3.187313e-08	8.568385e-22	8.941676e-10	...	4.120899e-11	2.329709e-12	0.0	5.694309e-13	1.016543e-11	1.165571e-07	1.803618e-09	0.0	0.0	5.091486e-10
			Ash for treatment, Re-processing of ash into clinker	0.0	0.0	0.000000e+00	0.0	0.0	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	...	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	0.000000e+00	0.0	0.0	0.000000e+00
			Aviation Gasoline	0.0	0.0	2.051568e-09	0.0	0.0	1.696496e-07	0.0	4.977002e-07	1.590425e-20	2.124223e-08	...	1.083025e-09	4.324300e-11	0.0	7.489337e-12	1.955008e-10	2.163469e-06	5.173984e-08	0.0	0.0	1.460579e-08
			BKB/Peat Briquettes	0.0	0.0	0.000000e+00	0.0	0.0	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	...	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.000000e+00	0.000000e+00	0.000000e+00	0.0	0.0	0.000000e+00
			Basic iron and steel and of ferro-alloys and first products thereof	0.0	0.0	4.762819e-05	0.0	0.0	8.046822e-04	0.0	6.932944e-03	3.392194e-17	2.298496e-04	...	1.400188e-05	9.223236e-08	0.0	1.438196e-07	2.289946e-06	4.614463e-03	2.453975e-04	0.0	0.0	6.927399e-05

10 rows × 49 columns

[37]:

value_added_concentration = db.calc_trades_concentration(
    indicator="total value added",
    item=SECTOR,
    scenario="2013",
    method="upstream",
    clusters="continent",
    clusters_direction="origin",
)

value_added_concentration

[37]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region
Europe	0.0	0.000000	0.669498	0.0	0.000000	0.218142	0.0	0.225551	0.311919	0.114916	...	0.21443	0.311919	0.0	0.410742	0.380046	0.311915	0.218150	0.000000	0.000000	0.218150
Oceania	0.0	0.449936	0.449936	0.0	0.449936	0.449936	0.0	0.449936	0.000000	0.000000	...	0.00000	0.000000	0.0	0.000000	0.449936	0.000000	0.000000	0.000000	0.449936	0.000000
America	0.0	0.000000	0.516330	0.0	0.516330	0.532473	0.0	0.318597	0.000000	0.000000	...	0.00000	0.000000	0.0	0.351588	0.000000	0.351588	0.000000	0.351566	0.532548	0.000000
Asia	0.0	0.176630	0.288928	0.0	0.000000	0.000000	0.0	0.203990	0.000000	0.000000	...	0.00000	0.218057	0.0	0.288905	0.267167	0.227852	0.288928	0.375378	0.292964	0.000000
Africa	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.330962	0.000000	0.000000	...	0.00000	0.000000	0.0	0.000000	0.000000	0.000000	0.321583	0.000000	0.000000	0.336972

5 rows × 49 columns

[20]:

value_added_exposure = db.calc_trades_exposure(
    indicator="total value added",
    exposed_to=focus_regions,
    item=SECTOR,
    scenario=base_scenario,
    method="upstream",
    clusters="continent",
    clusters_direction="origin",
)

value_added_exposure

[20]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region
Europe	0.0	0.000000	0.006872	0.0	0.000000	0.013652	0.0	0.013147	0.013172	0.010972	...	0.02102	0.013172	0.0	0.013260	0.010501	0.013172	0.013653	0.000000	0.000000	0.013653
Oceania	0.0	0.006377	0.006377	0.0	0.006377	0.006377	0.0	0.006377	0.000000	0.000000	...	0.00000	0.000000	0.0	0.000000	0.006377	0.000000	0.000000	0.000000	0.006377	0.000000
America	0.0	0.000000	0.003705	0.0	0.003705	0.005338	0.0	0.004143	0.000000	0.000000	...	0.00000	0.000000	0.0	0.004968	0.000000	0.004968	0.000000	0.004968	0.005338	0.000000
Asia	0.0	0.005992	0.004668	0.0	0.000000	0.000000	0.0	0.006026	0.000000	0.000000	...	0.00000	0.014183	0.0	0.004668	0.004845	0.008383	0.004668	0.004729	0.011143	0.000000
Africa	0.0	0.000000	0.000000	0.0	0.000000	0.000000	0.0	0.011062	0.000000	0.000000	...	0.00000	0.000000	0.0	0.000000	0.000000	0.000000	0.010961	0.000000	0.000000	0.005592

5 rows × 49 columns

Read the last two outputs together, not in isolation:

high concentration and high exposure: the embodied chain is dominated by a narrow set of contributors, and the selected Regions are a large part of that bottleneck
high concentration and low exposure: the chain is still concentrated, but the dominant contributors are mostly outside the Regions selected in exposed_to
low concentration and high exposure: the selected Regions matter a lot, but dependence is spread across several contributors rather than one single bottleneck
low concentration and low exposure: the chain is relatively diversified and not strongly tied to the selected Regions

So concentration is a question about the structure of the whole contributor base, while exposure is a question about dependence on a specific target.

7. Repeat the same logic for satellite accounts#

The same workflow can be reused for environmental extensions. First inspect the available labels, then choose the indicator that matches the footprint you want to track.

[38]:

db.search("CO2")

[38]:

{'Satellite account': ['CO2 - combustion - air',
  'CO2_bio - combustion - air',
  'CO2 - non combustion - Cement production - air',
  'CO2 - non combustion - Lime production - air',
  'CO2 - agriculture - peat decay - air',
  'CO2 - waste - biogenic - air',
  'CO2 - waste - fossil - air']}

[39]:

CO2_INDICATOR = "CO2 - combustion - air" # select one result of the search above

co2_upstream_content = db.calc_trades_content(
    indicator=CO2_INDICATOR,
    item=SECTOR,
    scenario=base_scenario,
    method="upstream",
    show_plot=True,
    auto_open=False,
    path=False,
    title=f"{CO2_INDICATOR} embodied in {SECTOR} trade",
)
co2_upstream_content.head(10)

INFO Resolver: resolving e for 2013.
INFO Resolver: trying e via formula build_iot_e_from_E_X.
INFO Resolver: resolved e via formula build_iot_e_from_E_X.
INFO Resolver: resolving f for 2013.
INFO Resolver: trying f via formula build_iot_f_from_e_w (compute_method=auto, runtime=inverse).
INFO Resolver: resolved f via formula build_iot_f_from_e_w (compute_method=auto, runtime=inverse).

[39]:

Region	AT	AU	BE	BG	BR	CA	CH	CN	CY	CZ	...	SK	TR	TW	US	WA	WE	WF	WL	WM	ZA
Region
AT	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0
AU	0.0	1.215395e+09	4874.435708	0.0	3.235317e+07	4.166857e+07	0.0	7.989606e+08	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	9046.622707	0.0	0.0	0.000000e+00	5.495651e+05	0.0
BE	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0
BG	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0
BR	0.0	0.000000e+00	18845.824215	0.0	1.272085e+08	2.547471e+04	0.0	1.011856e+06	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	1.198624e+03	0.000000e+00	0.0
CA	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	4.993165e+08	0.0	5.261423e+05	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	2.202058e+03	1.070593e+06	0.0
CH	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0
CN	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	7.642919e+09	0.0	0.000000	...	0.000000	0.0	0.0	7005.962304	436075.119212	0.0	0.0	7.730950e+06	0.000000e+00	0.0
CY	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	0.000000	...	0.000000	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0
CZ	0.0	0.000000e+00	0.000000	0.0	0.000000e+00	0.000000e+00	0.0	0.000000e+00	0.0	160947.623415	...	12999.551648	0.0	0.0	0.000000	0.000000	0.0	0.0	0.000000e+00	0.000000e+00	0.0

10 rows × 49 columns

8. Read the workflow as one complete supply-chain analysis#

A practical full-chain reading usually combines the methods above in this order:

calc_linkages(...): is the focal sector structurally important upstream, downstream, or both?
calc_trades(...): where are the relevant bilateral trade flows located?
calc_trades_content(...): how much value added or environmental pressure is embodied in those flows, and how much comes from direct production versus the wider supplier chain?
breakdown=True: which contributors actually generate that embodied content behind each trade pair?
calc_trades_concentration(...): is that embodied content diversified across many contributor Regions or dominated by a few bottlenecks?
calc_trades_exposure(...): what share of that embodied content depends on a specific Region or group of Regions?

That sequence turns MARIO from a parser into a full supply-chain analysis workflow: structure, geography, embodied indicators, hidden contributors, concentration, and dependence inside one EXIOBASE-backed database.

Download this notebook

Supply-chain analyses#

1. Define the EXIOBASE inputs and the focal supply chain#

2. Parse the first year and import the rest as scenarios#

3. Start from structural position: calc_linkages(...)#

4. Map the bilateral geography of the chain with calc_trades(...)#

5. Add embodied indicators with calc_trades_content(...)#

6. Explain the result with breakdown=True, concentration, and exposure#

7. Repeat the same logic for satellite accounts#

8. Read the workflow as one complete supply-chain analysis#

3. Start from structural position: `calc_linkages(...)`#

4. Map the bilateral geography of the chain with `calc_trades(...)`#

5. Add embodied indicators with `calc_trades_content(...)`#

6. Explain the result with `breakdown=True`, concentration, and exposure#