ModelFlowIb 2.93__tar.gz → 2.95__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (58) hide show
  1. {modelflowib-2.93 → modelflowib-2.95}/PKG-INFO +1 -1
  2. {modelflowib-2.93 → modelflowib-2.95}/model_parquet_mixin.py +29 -4
  3. {modelflowib-2.93 → modelflowib-2.95}/modelclass.py +19 -5
  4. {modelflowib-2.93 → modelflowib-2.95}/modelconstruct_estimation.py +62 -1
  5. {modelflowib-2.93 → modelflowib-2.95}/modelestimator_new.py +95 -7
  6. modelflowib-2.95/modelestimator_r.py +1492 -0
  7. {modelflowib-2.93 → modelflowib-2.95}/modelfactorize.py +240 -15
  8. {modelflowib-2.93 → modelflowib-2.95}/modelflowib.egg-info/PKG-INFO +1 -1
  9. {modelflowib-2.93 → modelflowib-2.95}/modelflowib.egg-info/SOURCES.txt +1 -0
  10. {modelflowib-2.93 → modelflowib-2.95}/modelflowib.egg-info/top_level.txt +1 -0
  11. {modelflowib-2.93 → modelflowib-2.95}/modelgrabgdx.py +42 -19
  12. {modelflowib-2.93 → modelflowib-2.95}/modelnewton.py +178 -14
  13. {modelflowib-2.93 → modelflowib-2.95}/modelreport.py +14 -10
  14. {modelflowib-2.93 → modelflowib-2.95}/modelsolver_ng.py +51 -22
  15. {modelflowib-2.93 → modelflowib-2.95}/pyproject.toml +3 -3
  16. {modelflowib-2.93 → modelflowib-2.95}/README.md +0 -0
  17. {modelflowib-2.93 → modelflowib-2.95}/license.md +0 -0
  18. {modelflowib-2.93 → modelflowib-2.95}/modelBLfunk.py +0 -0
  19. {modelflowib-2.93 → modelflowib-2.95}/model_Excel.py +0 -0
  20. {modelflowib-2.93 → modelflowib-2.95}/model_cvx.py +0 -0
  21. {modelflowib-2.93 → modelflowib-2.95}/model_dynare.py +0 -0
  22. {modelflowib-2.93 → modelflowib-2.95}/model_financial_stability.py +0 -0
  23. {modelflowib-2.93 → modelflowib-2.95}/model_latex.py +0 -0
  24. {modelflowib-2.93 → modelflowib-2.95}/model_latex_class.py +0 -0
  25. {modelflowib-2.93 → modelflowib-2.95}/modelcausality.py +0 -0
  26. {modelflowib-2.93 → modelflowib-2.95}/modelconstruct.py +0 -0
  27. {modelflowib-2.93 → modelflowib-2.95}/modeldash.py +0 -0
  28. {modelflowib-2.93 → modelflowib-2.95}/modeldashboot.py +0 -0
  29. {modelflowib-2.93 → modelflowib-2.95}/modeldashsidebar.py +0 -0
  30. {modelflowib-2.93 → modelflowib-2.95}/modeldekom.py +0 -0
  31. {modelflowib-2.93 → modelflowib-2.95}/modeldiff.py +0 -0
  32. {modelflowib-2.93 → modelflowib-2.95}/modeldisplay.py +0 -0
  33. {modelflowib-2.93 → modelflowib-2.95}/modelestimation.py +0 -0
  34. {modelflowib-2.93 → modelflowib-2.95}/modelflowib.egg-info/dependency_links.txt +0 -0
  35. {modelflowib-2.93 → modelflowib-2.95}/modelflowib.egg-info/requires.txt +0 -0
  36. {modelflowib-2.93 → modelflowib-2.95}/modelgrab.py +0 -0
  37. {modelflowib-2.93 → modelflowib-2.95}/modelgrabexcel.py +0 -0
  38. {modelflowib-2.93 → modelflowib-2.95}/modelgrabgekkobank.py +0 -0
  39. {modelflowib-2.93 → modelflowib-2.95}/modelgrabwf2.py +0 -0
  40. {modelflowib-2.93 → modelflowib-2.95}/modelhelp.py +0 -0
  41. {modelflowib-2.93 → modelflowib-2.95}/modelhtml.py +0 -0
  42. {modelflowib-2.93 → modelflowib-2.95}/modelinput_shiny.py +0 -0
  43. {modelflowib-2.93 → modelflowib-2.95}/modelinvert.py +0 -0
  44. {modelflowib-2.93 → modelflowib-2.95}/modeljupyter.py +0 -0
  45. {modelflowib-2.93 → modelflowib-2.95}/modeljupytermagic.py +0 -0
  46. {modelflowib-2.93 → modelflowib-2.95}/modelmacrograb.py +0 -0
  47. {modelflowib-2.93 → modelflowib-2.95}/modelmanipulation.py +0 -0
  48. {modelflowib-2.93 → modelflowib-2.95}/modelmf.py +0 -0
  49. {modelflowib-2.93 → modelflowib-2.95}/modelnet.py +0 -0
  50. {modelflowib-2.93 → modelflowib-2.95}/modelnormalize.py +0 -0
  51. {modelflowib-2.93 → modelflowib-2.95}/modelode.py +0 -0
  52. {modelflowib-2.93 → modelflowib-2.95}/modelpattern.py +0 -0
  53. {modelflowib-2.93 → modelflowib-2.95}/modeluserfunk.py +0 -0
  54. {modelflowib-2.93 → modelflowib-2.95}/modelvis.py +0 -0
  55. {modelflowib-2.93 → modelflowib-2.95}/modelwidget.py +0 -0
  56. {modelflowib-2.93 → modelflowib-2.95}/modelwidget_core.py +0 -0
  57. {modelflowib-2.93 → modelflowib-2.95}/modelwidget_input.py +0 -0
  58. {modelflowib-2.93 → modelflowib-2.95}/setup.cfg +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: modelflowib
3
- Version: 2.93
3
+ Version: 2.95
4
4
  Summary: A tool to solve and manage dynamic economic and financial models
5
5
  Author-email: Ib Hansen <Ib.Hansen.Iv@gmail.com>
6
6
  License-Expression: MIT
@@ -84,6 +84,24 @@ def _period_series_to_json(series):
84
84
  return json.dumps(result_dict)
85
85
 
86
86
 
87
+ def storable_options(options):
88
+ """The sticky solve options (oldkwargs) that can be written as JSON. A function with a ``spec`` dict (the
89
+ factorizers from modelfactorize.normal/pardiso) is stored as that dict. Others, such as a user function given
90
+ as factorizer=, are left out with a note: the loaded model then uses the default for them."""
91
+ keep, dropped = {}, []
92
+ for k, v in (options or {}).items():
93
+ if callable(v) and isinstance(getattr(v, 'spec', None), dict):
94
+ v = v.spec
95
+ try:
96
+ json.dumps(v)
97
+ keep[k] = v
98
+ except (TypeError, ValueError):
99
+ dropped.append(k)
100
+ if dropped:
101
+ print(f'modeldump: the solve option(s) {", ".join(dropped)} can not be saved and are left out')
102
+ return keep
103
+
104
+
87
105
  def _df_to_feather_bytes(df):
88
106
  """Serialize a DataFrame to in-memory Feather/Arrow IPC bytes.
89
107
 
@@ -164,7 +182,8 @@ class Parquet_Mixin:
164
182
 
165
183
  # ── dump ──────────────────────────────────────────────────────────────
166
184
 
167
- def modeldump(self, file_path='', keep=False, large=False, compact=False, derivatives=False, **kwargs):
185
+ def modeldump(self, file_path='', keep=False, large=False, compact=False, derivatives=False, savebase=False,
186
+ **kwargs):
168
187
  """Dump model to disk.
169
188
 
170
189
  Parameters
@@ -187,6 +206,11 @@ class Parquet_Mixin:
187
206
  session takes them from the model instead of finding them again, which is
188
207
  the slow part of its first solve for a large model. They are used only if
189
208
  the equations and the solver options are the same.
209
+ savebase : bool
210
+ If True, store the basedf instead of the lastdf (for instance the baseline after an
211
+ experiment has been run). Only one dataframe is stored, so the file stays small; kept
212
+ solutions can be added with ``keep=True``. ``modelload`` gives it back as the dataframe
213
+ of the file, as always.
190
214
 
191
215
  A model instance without a lastdf holds no dataframe, so the feather
192
216
  format has nothing to be fast about. Such a model is always written as a
@@ -200,7 +224,8 @@ class Parquet_Mixin:
200
224
 
201
225
  if not large or nodata:
202
226
  # ── original JSON/gzip path ──────────────────────────────────
203
- return super().modeldump(file_path=file_path, keep=keep, derivatives=derivatives, **kwargs)
227
+ return super().modeldump(file_path=file_path, keep=keep, derivatives=derivatives, savebase=savebase,
228
+ **kwargs)
204
229
 
205
230
  # ── fast feather/zip path ────────────────────────────────────────
206
231
  pathname = Path(file_path)
@@ -222,7 +247,7 @@ class Parquet_Mixin:
222
247
  'format': 'feather_zip',
223
248
  'frml': self.equations,
224
249
  'modelname': self.name,
225
- 'oldkwargs': self.__dict__.get('oldkwargs', {}), # a model never solved has none
250
+ 'oldkwargs': storable_options(self.__dict__.get('oldkwargs', {})), # a model never solved has none
226
251
  'run_on_load': self.__dict__.get('run_on_load', False),
227
252
  'straight': bool(self.__dict__.get('straight', False)), # loads the same way again
228
253
  'diff_expressions': self.__dict__.get('diff_expressions', {}) if derivatives else {},
@@ -261,7 +286,7 @@ class Parquet_Mixin:
261
286
  # current_per
262
287
  zf.writestr('current_per.json', current_per_json)
263
288
  # lastdf → feather
264
- zf.writestr('lastdf.feather', _df_to_feather_bytes(_maybe_compact(self.dump_df)))
289
+ zf.writestr('lastdf.feather', _df_to_feather_bytes(_maybe_compact(self.dump_frame(savebase))))
265
290
  # keep_solutions → one feather each, indexed by position
266
291
  if keep:
267
292
  for i, (name, df) in enumerate(self.keep_solutions.items()):
@@ -6865,11 +6865,24 @@ class Json_Mixin():
6865
6865
 
6866
6866
  return pathname, dump_data
6867
6867
 
6868
- def modeldump_base(self, outfile='',keep=False,data=True,derivatives=False):
6868
+ def dump_frame(self, savebase=False):
6869
+ '''The dataframe a dump stores: dump_df (normally the lastdf), or with savebase=True the basedf instead.
6870
+ Only one dataframe is stored, so the file stays small; kept solutions can be added with keep=True.'''
6871
+ if savebase:
6872
+ base = self.__dict__.get('basedf')
6873
+ if isinstance(base, pd.DataFrame):
6874
+ return base
6875
+ print('modeldump: savebase=True, but the model instance has no basedf - the lastdf is dumped')
6876
+ return self.dump_df
6877
+
6878
+ def modeldump_base(self, outfile='',keep=False,data=True,derivatives=False,savebase=False):
6869
6879
  '''Dumps a model and its lastdf to a json file
6870
6880
 
6871
6881
  if keep=True the model.keep_solutions will alse be dumped
6872
6882
 
6883
+ if savebase=True the basedf is dumped instead of the lastdf (for instance the baseline after an
6884
+ experiment has been run). modelload gives it back as the dataframe of the file, as always
6885
+
6873
6886
  if derivatives=True the expressions for the derivatives, which a Newton solver has found
6874
6887
  (model.diff_expressions), are dumped too. A Newton solver in a later session then takes them
6875
6888
  from the model instead of finding them again - the slow part of its first solve for a large model
@@ -6925,14 +6938,15 @@ class Json_Mixin():
6925
6938
  current_per_json = None
6926
6939
 
6927
6940
 
6941
+ from model_parquet_mixin import storable_options
6928
6942
  dumpjson = {
6929
6943
  'version': '1.10',
6930
6944
  'has_data': dump_data,
6931
6945
  'frml': self.equations,
6932
- 'lastdf': self.dump_df.to_json(double_precision=15) if dump_data else None,
6946
+ 'lastdf': self.dump_frame(savebase).to_json(double_precision=15) if dump_data else None,
6933
6947
  'current_per': current_per_json ,
6934
6948
  'modelname': self.name,
6935
- 'oldkwargs': self.__dict__.get('oldkwargs',{}),
6949
+ 'oldkwargs': storable_options(self.__dict__.get('oldkwargs',{})),
6936
6950
  'run_on_load': self.__dict__.get('run_on_load', False),
6937
6951
  'straight': bool(self.__dict__.get('straight', False)), # loads the same way again
6938
6952
  'diff_expressions': self.__dict__.get('diff_expressions', {}) if derivatives else {},
@@ -7134,12 +7148,12 @@ class Json_Mixin():
7134
7148
 
7135
7149
  class Zip_Mixin():
7136
7150
  '''This experimental class zips a dumped file '''
7137
- def modeldump(self, file_path='',keep=False,zip_file=True,data=True,derivatives=False):
7151
+ def modeldump(self, file_path='',keep=False,zip_file=True,data=True,derivatives=False,savebase=False):
7138
7152
  pathname, dump_data = self._dump_pathname(file_path, self.has_data, data=data)
7139
7153
 
7140
7154
  pathname.parent.mkdir(parents=True, exist_ok=True)
7141
7155
 
7142
- json_string = self.modeldump_base(keep=keep,data=dump_data,derivatives=derivatives)
7156
+ json_string = self.modeldump_base(keep=keep,data=dump_data,derivatives=derivatives,savebase=savebase)
7143
7157
 
7144
7158
 
7145
7159
  if zip_file:
@@ -1600,6 +1600,9 @@ BUILTIN_ESTIMATORS = {
1600
1600
  'ols': 'Estimate_ols',
1601
1601
  'nls_lmfit': 'Estimate_nls_lmfit',
1602
1602
  'nls_eviews': 'Estimate_nls_eviews',
1603
+ # R as the estimation engine (modelestimator_r); method lm / nls by default
1604
+ 'r_lm': 'Estimate_r_lm',
1605
+ 'r_nls': 'Estimate_r_nls',
1603
1606
  }
1604
1607
 
1605
1608
 
@@ -1650,6 +1653,11 @@ def _get_estimator_class(estimator_name, estimator_classes: Optional[dict] = Non
1650
1653
  "Equations with <estimator=...> require modelestimator_new to be importable."
1651
1654
  ) from exc
1652
1655
 
1656
+ if class_names[name].startswith('Estimate_r_'):
1657
+ # The R backends live in their own module so modelestimator_new
1658
+ # does not depend on them.
1659
+ import modelestimator_r as me
1660
+
1653
1661
  try:
1654
1662
  return getattr(me, class_names[name])
1655
1663
  except AttributeError as exc:
@@ -2366,7 +2374,8 @@ class Makemodel(BaseExplode):
2366
2374
  self.markdown_model = (mfmod_list_to_codeblock(self.original_statements))
2367
2375
 
2368
2376
  elif self.type_input == 'single':
2369
- extracted_frml = self.original_statements.replace('\n', '')
2377
+ # join the lines with a space, so a tag ending a line stays apart from the equation
2378
+ extracted_frml = ' '.join(line.strip() for line in self.original_statements.splitlines()).strip()
2370
2379
  self.markdown_model = ''
2371
2380
 
2372
2381
  else:
@@ -2685,6 +2694,46 @@ class Makemodel(BaseExplode):
2685
2694
  cols[name] = ser
2686
2695
  return pd.DataFrame(cols)
2687
2696
 
2697
+ def actual_fitted(self, level: bool = True):
2698
+ """Actual, fitted and residuals of all estimated equations as one DataFrame.
2699
+
2700
+ The columns have two levels: the endogenous variable, and ``Actual``,
2701
+ ``Fitted``, ``Residuals``. Take one equation with
2702
+ ``mm.actual_fitted()['CON']``.
2703
+
2704
+ level : bool, default True
2705
+ True: the endogenous variable itself. When the left-hand side is
2706
+ ``DLOG(CON)`` this is ``CON``, its one-step fitted value (the
2707
+ estimated equation solved for ``CON``, lagged values are the
2708
+ actual ones) and the residual in the units of ``CON``.
2709
+ False: the left-hand side as it was estimated (``DLOG(CON)``).
2710
+
2711
+ Each equation covers its own estimation sample, so rows outside it
2712
+ are NaN. An equation which can't be solved for its variable is left
2713
+ out with ``level=True``.
2714
+
2715
+ Per-equation access: ``rec['estimator_object'].af_level_df`` (or
2716
+ ``.af_df`` and ``.residuals_df``) on the entries of
2717
+ :attr:`estimation_records`; ``modelestimator_new.plot_actual_vs_fitted
2718
+ (est, level=True)`` draws the chart of the report.
2719
+ """
2720
+ import pandas as pd
2721
+ frames = {}
2722
+ for rec in self.estimation_records:
2723
+ est = rec.get('estimator_object')
2724
+ if not _is_estimator_backend_instance(est):
2725
+ continue
2726
+ frame = est.af_level_df if level else est.af_df.join(est.residuals_df)
2727
+ if frame is None:
2728
+ continue
2729
+ name = getattr(est, 'endo_var', '') or rec.get('frmlname', '')
2730
+ if level: # the variable is the outer column level
2731
+ frame = frame.rename(columns=lambda c: c.removeprefix(f'{est.endo_var}__'))
2732
+ frames[name] = frame
2733
+ if not frames:
2734
+ return pd.DataFrame()
2735
+ return pd.concat(frames, axis=1)
2736
+
2688
2737
  @property
2689
2738
  def markdown_with_estimation(self) -> str:
2690
2739
  """Original Markdown input with compact estimation tables inserted.
@@ -3085,6 +3134,18 @@ class Listmodels(BaseExplode):
3085
3134
  return pd.DataFrame()
3086
3135
  return pd.concat(frames, axis=1)
3087
3136
 
3137
+ def actual_fitted(self, level: bool = True):
3138
+ """Actual, fitted and residuals of all member Makemodels, combined column-wise.
3139
+
3140
+ See :meth:`Makemodel.actual_fitted` for the layout.
3141
+ """
3142
+ import pandas as pd
3143
+ frames = [mex.actual_fitted(level=level) for mex in self.makemodels]
3144
+ frames = [f for f in frames if not f.empty]
3145
+ if not frames:
3146
+ return pd.DataFrame()
3147
+ return pd.concat(frames, axis=1)
3148
+
3088
3149
  @property
3089
3150
  def markdown_with_estimation(self) -> str:
3090
3151
  """Concatenate member Makemodel markdown-with-estimation strings."""
@@ -1378,6 +1378,67 @@ class EstimatorBackend(Eq_parent, ABC):
1378
1378
  .rename(columns=lambda s: s.lower().capitalize())
1379
1379
  )
1380
1380
 
1381
+ # ---- A/F and residuals of the endogenous variable itself ---------------
1382
+
1383
+ @cached_property
1384
+ def lhs_is_transformed(self) -> bool:
1385
+ """True when the left-hand side is a transformation of the endogenous
1386
+ variable (``DLOG(CON)``, ``DIFF(CON)``, ``LOG(CON/P)`` ...), so
1387
+ :attr:`af_df` is not in the units of the variable itself."""
1388
+ lhs = self.org_eq_clean.split("=", 1)[0]
1389
+ return "".join(lhs.split()).upper() != self.endo_var.upper()
1390
+
1391
+ @cached_property
1392
+ def af_level_df(self) -> Optional[pd.DataFrame]:
1393
+ """Actual, fitted and residuals of the endogenous variable itself.
1394
+
1395
+ :attr:`af_df` and :attr:`residuals_df` are in the units of the
1396
+ left-hand side as it was estimated, e.g. ``DLOG(CON)``. Here the
1397
+ estimated equation is solved for the variable (``CON``):
1398
+
1399
+ * ``Actual`` the variable,
1400
+ * ``Fitted`` the normalized equation with the estimated
1401
+ coefficients, evaluated on the data. Lagged values of the variable
1402
+ are the actual ones, so this is the one-step (static) fit, not a
1403
+ dynamic simulation,
1404
+ * ``Residuals`` ``Actual - Fitted``, in the units of the variable.
1405
+
1406
+ The columns are prefixed with the variable name and ``__``
1407
+ (``CON__Actual``, ``CON__Fitted``, ``CON__Residuals``), so frames of
1408
+ several equations can be joined side by side.
1409
+
1410
+ When the left-hand side is the variable itself the frame holds the
1411
+ same numbers as :attr:`af_df` and :attr:`residuals_df`. ``None`` (and
1412
+ a message) when the equation can't be solved for the variable.
1413
+ """
1414
+ prefix = f"{self.endo_var}__"
1415
+ if not self.lhs_is_transformed:
1416
+ return self.af_df.join(self.residuals_df).add_prefix(prefix)
1417
+ start, end = self.smpl
1418
+ try:
1419
+ normalized = nz.normal(
1420
+ self.org_eq_baked, add_add_factor=False
1421
+ ).normalized
1422
+ lhs, rhs = normalized.split("=", 1)
1423
+ if lhs.strip().upper() != self.endo_var.upper():
1424
+ raise ValueError(f"normalized to {lhs.strip()!r}")
1425
+ res = self.eq_var_df.mfcalc(
1426
+ f"<{start},{end}> FITTED_LEVEL = {rhs}"
1427
+ )
1428
+ out = pd.DataFrame({
1429
+ "Actual": res.loc[start:end, self.endo_var],
1430
+ "Fitted": res.loc[start:end, "FITTED_LEVEL"],
1431
+ })
1432
+ except Exception as e:
1433
+ print(
1434
+ f"{type(self).__name__}({self.endo_var}): no actual/fitted "
1435
+ f"for the variable itself, the equation could not be solved "
1436
+ f"for it ({type(e).__name__}: {str(e)[:100]})"
1437
+ )
1438
+ return None
1439
+ out["Residuals"] = out["Actual"] - out["Fitted"]
1440
+ return out.add_prefix(prefix)
1441
+
1381
1442
  @cached_property
1382
1443
  def estimation_smpl(self) -> Tuple[Any, Any]:
1383
1444
  """The usable sample, derived from non-missing residuals.
@@ -2088,6 +2149,11 @@ class LSResult:
2088
2149
  title_html = self._build_title_html()
2089
2150
  summary_html = self.olsmodel._render_summary_html()
2090
2151
  plot_html = self._build_plot_html(plot_format)
2152
+ # the same for the endogenous variable itself, when the left-hand
2153
+ # side is a transformation of it (DLOG(CON) -> CON)
2154
+ if (self.olsmodel.lhs_is_transformed
2155
+ and self.olsmodel.af_level_df is not None):
2156
+ plot_html += self._build_plot_html(plot_format, level=True)
2091
2157
  return title_html + summary_html + plot_html
2092
2158
 
2093
2159
  def show(self, plot_format: str = "svg") -> None:
@@ -2124,17 +2190,21 @@ class LSResult:
2124
2190
  f"<pre><code>{normalized}</code></pre></p>"
2125
2191
  )
2126
2192
 
2127
- def _build_plot_html(self, plot_format: str) -> str:
2193
+ def _build_plot_html(self, plot_format: str, level: bool = False) -> str:
2128
2194
  buf = BytesIO()
2129
2195
  plot_actual_vs_fitted(
2130
2196
  self.olsmodel, save_to=buf, format=plot_format,
2131
- figsize=(8, 5), dpi=150,
2197
+ figsize=(8, 5), dpi=150, level=level,
2132
2198
  )
2133
2199
  buf.seek(0)
2134
2200
  b64 = base64.b64encode(buf.read()).decode("utf-8")
2135
2201
  mime = "svg+xml" if plot_format == "svg" else "png"
2202
+ heading = (
2203
+ f"Actual vs Fitted Plot: {self.olsmodel.endo_var} itself "
2204
+ "(one-step fit)" if level else "Actual vs Fitted Plot"
2205
+ )
2136
2206
  return (
2137
- "<h3>Actual vs Fitted Plot</h3>"
2207
+ f"<h3>{heading}</h3>"
2138
2208
  f'<img style="max-width:100%; height:auto;" '
2139
2209
  f'src="data:image/{mime};base64,{b64}" />'
2140
2210
  )
@@ -2440,28 +2510,46 @@ def plot_actual_vs_fitted(
2440
2510
  format: str = "png",
2441
2511
  figsize: Tuple[float, float] = (7, 3),
2442
2512
  dpi: int = 150,
2513
+ level: bool = False,
2443
2514
  ):
2444
2515
  """Plot Actual vs Fitted (top panel) and Residuals (bottom panel).
2445
2516
 
2517
+ ``level=False`` plots the left-hand side as it was estimated (``af_df``,
2518
+ ``residuals_df``), ``level=True`` the endogenous variable itself
2519
+ (``af_level_df``): ``CON`` when the left-hand side is ``DLOG(CON)``.
2520
+
2446
2521
  If ``save_to`` is a file-like object, the plot is saved there and the
2447
2522
  figure is closed; otherwise it is shown interactively.
2448
2523
  """
2524
+ if level:
2525
+ if emodel.af_level_df is None:
2526
+ raise ValueError(
2527
+ f"{emodel.endo_var}: no actual/fitted for the variable itself"
2528
+ )
2529
+ prefix = f"{emodel.endo_var}__"
2530
+ af_df = residuals_df = emodel.af_level_df.rename(
2531
+ columns=lambda c: c.removeprefix(prefix))
2532
+ title = f"Actual vs Fitted: {emodel.endo_var} itself (one-step fit)"
2533
+ else:
2534
+ af_df, residuals_df = emodel.af_df, emodel.residuals_df
2535
+ title = f"Actual vs Fitted: {emodel.endo_var}"
2536
+
2449
2537
  fig = plt.figure(figsize=figsize, dpi=dpi)
2450
2538
  gs = GridSpec(3, 1, height_ratios=[2, 0.1, 1])
2451
2539
 
2452
2540
  ax1 = fig.add_subplot(gs[0])
2453
2541
  ax2 = fig.add_subplot(gs[2], sharex=ax1)
2454
2542
 
2455
- ax1.plot(emodel.af_df.index, emodel.af_df["Actual"],
2543
+ ax1.plot(af_df.index, af_df["Actual"],
2456
2544
  label="Actual", linewidth=2)
2457
- ax1.plot(emodel.af_df.index, emodel.af_df["Fitted"],
2545
+ ax1.plot(af_df.index, af_df["Fitted"],
2458
2546
  label="Fitted", linestyle="--")
2459
- ax1.set_title(f"Actual vs Fitted: {emodel.endo_var}")
2547
+ ax1.set_title(title)
2460
2548
  ax1.set_ylabel(emodel.endo_var)
2461
2549
  ax1.grid(True)
2462
2550
  ax1.legend()
2463
2551
 
2464
- ax2.plot(emodel.residuals_df.index, emodel.residuals_df["Residuals"],
2552
+ ax2.plot(residuals_df.index, residuals_df["Residuals"],
2465
2553
  color="gray")
2466
2554
  ax2.axhline(0, color="red", linestyle="--", linewidth=1)
2467
2555
  ax2.set_title("Residuals")