Visualization workflows

This notebook shows the new unified db.plot(...) workflow on the packaged MARIO test databases.

The same plotting engine supports both a guided path for quick inspection and an advanced path with explicit Plotly Express mappings, including custom palettes.

import mario

Load the packaged test databases

These are the same small MARIO-readable tables used elsewhere in the documentation and test suite, so the examples are reproducible without external downloads.

iot = mario.load_test("IOT")
sut = mario.load_test("SUT")

INFO Parser: excel reading IOT flows from /Users/lorenzorinaldi/Documents/GitHub/MARIO/mario/test/tables/test_IOT_standard.xlsx.
INFO Parser: state payload ready with 6 canonical blocks.
INFO Parser: excel state ready for IOT.
INFO Metadata: initialized.
INFO Parser: excel reading SUT flows from /Users/lorenzorinaldi/Documents/GitHub/MARIO/mario/test/tables/test_SUT_standard.xlsx.
INFO Parser: state payload ready with 10 canonical blocks.
INFO Parser: excel state ready for SUT.
INFO Metadata: initialized.

Quick path: plot one matrix with a preset

For a first look at a matrix, pass the matrix name and let MARIO choose a useful Plotly Express mapping with a preset such as "overview".

fig = iot.plot(
    matrix="Z",
    preset="overview",
    auto_open=False,
    path=False,
)
fig

Inspect the dataframe actually plotted

Set return_data=True when you want both the Plotly figure and the dataframe passed to Plotly Express after filtering, aggregation and top-n trimming.

fig, plotted = iot.plot(
    matrix="Z",
    preset=None,
    kind="bar",
    x="Sector_from",
    color="Region_to",
    auto_open=False,
    path=False,
    return_data=True,
)
plotted.head()

	Sector_from	Region_to	Value
0	Agriculture	Reg1	7.367798e+06
1	Agriculture	Reg2	1.048720e+05
2	Industry	Reg1	3.487216e+07
3	Industry	Reg2	6.248561e+05
4	Services	Reg1	4.228452e+07

Advanced path: explicit Plotly Express mappings

More experienced users can control the chart directly through x, color, facet_col, animation_frame, agg, top_n and the other familiar mappings.

fig = iot.plot(
    matrix="Y",
    kind="bar",
    preset=None,
    x="Sector_from",
    color="Consumption category_to",
    facet_col="Region_to",
    agg="sum",
    top_n=6,
    auto_open=False,
    path=False,
)
fig

Filter before plotting

Use filters={...} to keep only the levels you want before MARIO aggregates the data.

fig = iot.plot(
    matrix="Z",
    kind="bar",
    preset=None,
    x="Sector_from",
    color="Sector_to",
    filters={"Region_from": ["Reg1"], "Region_to": ["Reg1"]},
    top_n=4,
    auto_open=False,
    path=False,
)
fig

SUT-specific views

The same engine works on split SUT blocks such as U and S, so you can plot activity and commodity dimensions directly.

fig = sut.plot(
    matrix="U",
    kind="bar",
    preset=None,
    x="Commodity_from",
    color="Activity_to",
    facet_col="Region_to",
    top_n=4,
    auto_open=False,
    path=False,
)
fig

Plot region-by-region trades for one item

db.calc_trades(...) can aggregate intermediate and final trade into one region-by-region matrix. Use total=True to add row and column totals.

trade_matrix = iot.calc_trades(
    "Agriculture",
    total=True,
)
trade_matrix

Region	Reg1	Reg2	Total
Region
Reg1	9.897401e+06	74768.097334	9.972169e+06
Reg2	8.933711e+03	70915.947332	7.984966e+04
Total	9.906335e+06	145684.044667	1.005202e+07

Set aggregate=False and show_plot=True to keep intermediate and final trade separate and produce a heatmap with one panel per component. The plot is displayed as a side effect; the return value remains the trade matrix.

trade_components = iot.calc_trades(
    "Agriculture",
    aggregate=False,
    show_plot=True,
    path=False,
    auto_open=False,
)
trade_components

	Intermediate		Final
Region	Reg1	Reg2	Reg1	Reg2
Region
Reg1	7.363942e+06	63312.752520	2.533459e+06	11455.344815
Reg2	3.855907e+03	41559.247333	5.077805e+03	29356.699999

Plot a prepared dataframe

db.plot(...) can also accept a custom dataframe when the quantity to visualize is not one raw MARIO matrix. Here we reuse sectoral GDP.

gdp = iot.GDP(total=False).reset_index()
gdp

	Region	Sector	GDP
0	Reg1	Agriculture	6.016424e+06
1	Reg1	Industry	1.497569e+07
2	Reg1	Services	6.206624e+07
3	Reg2	Agriculture	4.539138e+04
4	Reg2	Industry	3.520592e+05
5	Reg2	Services	1.428250e+06

fig = iot.plot(
    data=gdp,
    kind="treemap",
    preset=None,
    y="GDP",
    path_columns=["Sector"],
    auto_open=False,
    path=False,
)
fig

Change the default palette

Use mario.set_palette(...) to change the default discrete color sequence used by db.plot(...). MARIO ships built-in palettes such as mario, Plotly, Safe, Vivid, Pastel, Alphabet, and McKinsey.

mario.set_palette("Safe")

fig = iot.plot(
    matrix="Y",
    kind="bar",
    preset=None,
    x="Sector_from",
    color="Consumption category_to",
    facet_col="Region_to",
    agg="sum",
    top_n=6,
    auto_open=False,
    path=False,
)
fig

You can also pass an explicit list of colors. The palette is session-wide, so reset it when you want to go back to the MARIO defaults.

mario.set_palette("mario")

fig = iot.plot(
    matrix="Z",
    kind="sunburst",
    preset=None,
    path_columns=["Region_from", "Sector_from"],
    top_n=8,
    auto_open=False,
    path=False,
)
fig

Download this notebook