Scripting wolfhece from a VSCode Jupyter Notebook

This notebook shows the recommended way to drive a live WolfMapViewer interactively from VSCode — no bridge, no separate server, no extra dependencies beyond ipykernel (already used by VSCode itself).

How it works

IPython has built-in support for wx event-loop integration via the %gui wx magic. When active, IPython drives the wx event loop itself using a timer-based mechanism — you must not call app.MainLoop() because IPython handles it.

%gui wx        ← tells IPython to manage the wx event loop
                   run once, alone in its own cell, before any wx code

Each notebook cell returns immediately; wx windows are live and keep processing events while you run subsequent cells.

Two entry points

Class	When to use
WolfMapViewer	Minimal viewer only — no extra data, fastest start
MapManager	Full production stack (hydrometry, PICC, cadaster, WMS, …)

Sections A and B below cover each option.

---

Requirements — wolfhece installed in the active kernel’s environment. The kernel is the same Python that runs in the VSCode terminal.

# ── Step 1 (mandatory, run this cell first, alone) ───────────────────────────
# Activate wx event-loop integration.  After this cell, wx windows are live
# and the notebook remains interactive simultaneously.
%gui wx

# ── Step 2 ───────────────────────────────────────────────────────────────────
# Create the wx.App singleton.  False = do not redirect stdout/stderr.
# Do NOT call app.MainLoop() — %gui wx handles the event loop.
# import wx
# app = wx.App(False)

---

Section A — Minimal viewer (WolfMapViewer)

Use this when you only need the map canvas, without the full data stack (no hydrometry GUI, no PICC/cadaster layers, etc.).

from wolfhece.PyDraw import WolfMapViewer

viewer = WolfMapViewer(None, title='wolfhece — notebook session')
viewer.Show()

INFO:root:Importing wolfhece modules
INFO:root:wolfhece modules imported

False

# Cells below execute while the viewer is open.

# Current viewport bounds:
print(f"x: [{viewer.xmin:.1f}, {viewer.xmax:.1f}]")
print(f"y: [{viewer.ymin:.1f}, {viewer.ymax:.1f}]")

x: [-2.0, 42.0]
y: [-0.9, 40.9]

# Load a raster array and display it.
# Adapt the path to your own file.
from pathlib import Path

array_file = Path(r'C:\path\to\my_array.bin')   # ← adapt

if array_file.exists():
    viewer.add_object(which='array', filename=str(array_file), id='DEM', ToCheck=True)
    viewer.findminmax(True)
    viewer.Autoscale(False)
    viewer.Paint()
else:
    print(f"File not found: {array_file}")

File not found: C:\path\to\my_array.bin

# Direct wx calls are safe — we are on the main thread.
viewer.set_statusbar_text('Hello from the notebook!')
viewer.Paint()

---

Section B — Full production stack (MapManager)

MapManager wraps a WolfMapViewer and automatically loads the standard data layers: SPW hydrometry stations, PICC parcels, cadaster, land maps, WMS tiles, coordinate grid, and optional HECE database items.

Key points

• Pass splash=False to suppress the splash screen (not needed in a notebook).

• The viewer itself is at manager.mapviewer.

• MapManager is itself a (hidden) wx.Frame — you do not need to call manager.Show(). The viewer frame shows itself.

# Close the minimal viewer from Section A first if it is still open.
# viewer.Close()

from wolfhece.PyGui import MapManager

# splash=False  → skip the WolfLauncher splash screen
manager = MapManager(splash=False)

# Convenient shorthand
v = manager.mapviewer

INFO:root:Importing wolfhece modules
INFO:root:wolfhece modules imported
INFO:root:MapManager
INFO:root:MapManager - MapViewer created
INFO:root:MapManager - Data directories created
ERROR:root:Error in hydrometry init: 'numpy.ndarray' object is not callable
INFO:root:MapManager - hydrometry_wolfgui created
INFO:root:MapManager - Picc_data created
INFO:root:MapManager - Cadaster_data created
INFO:root:MapManager - PlansTerrier created
INFO:root:MapManager - Ouvrages Ponts created
INFO:root:MapManager - Ouvrages Seuils created
INFO:root:MapManager - Enquetes created
INFO:root:MapManager - Particularites created
INFO:root:MapManager - PlansTerrier created
INFO:root:MapManager - Menu Landuse/Landcover created

# Verify what was loaded automatically.
from wolfhece.PyDraw import draw_type

arrays  = v.get_list_ids(drawing_type=draw_type.ARRAYS,  checked_state=None)
vectors = v.get_list_ids(drawing_type=draw_type.VECTORS, checked_state=None)

print("Arrays  :", arrays  or '(none)')
print("Vectors :", vectors or '(none)')

Arrays  : (none)
Vectors : ['grid']

# Add a raster, fit the view and repaint.
from pathlib import Path

array_file = Path(r'C:\path\to\my_array.bin')   # ← adapt

if array_file.exists():
    v.add_object(which='array', filename=str(array_file), id='DEM', ToCheck=True)
    v.findminmax(True)
    v.Autoscale(False)
    v.Paint()
else:
    print(f"File not found: {array_file}")

File not found: C:\path\to\my_array.bin

# Inspect the active array (selected in the tree on the left).
a = v.active_array
if a is not None:
    print(f"Active array : {a.idx}")
    print(f"  Shape      : {a.array.shape}")
    print(f"  Min / Max  : {float(a.array.min()):.3f} / {float(a.array.max()):.3f}")
else:
    print('No active array — click one in the tree.')

No active array — click one in the tree.

# Access the SPW hydrometry sub-manager loaded by MapManager.
hydro = manager.SPWhydrometry
print(type(hydro))
# hydro.get_stations()   ← example method (check the API)

<class 'wolfhece.hydrometry.kiwis_wolfgui.hydrometry_wolfgui'>

---

Section C — Useful scripting patterns

A collection of common operations once the viewer is live.

# Work with the active array as a NumPy masked array.
import numpy as np

a = v.active_array
if a is not None:
    data = a.array          # np.ma.MaskedArray, shape (cols, rows)
    valid = data[~data.mask]
    print(f"Valid cells : {valid.size:,}")
    print(f"Mean        : {valid.mean():.3f}")
    print(f"Std         : {valid.std():.3f}")

# Save the current canvas to PNG.
from pathlib import Path

out = Path.home() / 'wolf_snapshot.png'
v.save_canvasogl(fn=str(out))
print(f'Saved → {out}')

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[9], line 5
      2 from pathlib import Path
      4 out = Path.home() / 'wolf_snapshot.png'
----> 5 v.save_canvas(str(out))
      6 print(f'Saved → {out}')

AttributeError: 'WolfMapViewer' object has no attribute 'save_canvas'

# Programmatic zoom to a bounding box (Lambert 72 / 2008 coords).
# v.zoom_on_bounds(xmin, ymin, xmax, ymax)
# v.Autoscale(False)
# v.Paint()

---

Section D — Comparison with the standalone app (wolf.py)

# wolf.py  (standalone)
ex = wx.App()
mywolf = MapManager()       # splash=True by default
ex.MainLoop()               # blocks until the window closes

# notebook  (this file)
%gui wx                     # IPython manages the event loop
app = wx.App(False)
manager = MapManager(splash=False)   # no MainLoop() call
v = manager.mapviewer        # keep running cells while the window is live

The only differences are %gui wx, splash=False, and the absence of MainLoop(). Everything else — objects, methods, data — is identical.

---

Section E — Closing

Close the viewer window normally, or from a cell:

# v.Close()      # closes the viewer frame
# manager.Close()  # also closes the manager frame (log window, etc.)