def _create_tabs(self):
"Creates a tab for each variable"
self.lis = self.classl if self.is_ordered else list(
permutations(self.classl, 2))
if self._is_tab:
self._boxes = len(self.df_list)
self._cols = math.ceil(math.sqrt(self._boxes))
self._rows = math.ceil(self._boxes / self._cols)
self.tbnames = list(
self.df_list[0].columns
)[:-1] if self.varlist is None else self.varlist
else:
vals = self.interp.most_confused()
self._ranges = []
self.tbnames = []
self._boxes = int(
input(
'Please enter a value for `k`, or the top images you will see: '
))
for x in iter(vals):
for y in range(len(self.lis)):
if x[0:2] == self.lis[y]:
self._ranges.append(x[2])
self.tbnames.append(str(x[0] + ' | ' + x[1]))
self.tb = widgets.TabBar(self.tbnames)
self._populate_tabs()
def __init__(self,
learn,
dl=None,
cut_off=100,
is_ordered=False,
classlist=[],
varlist=None,
figsize=(12, 12),
**kwargs):
dl = learn.dls[1] if dl is None else dl
interp = ClassificationInterpretation.from_learner(learn, dl=dl)
combs = classlist if is_ordered else list(
itertools.permutations(classlist, 2))
figsize = figsize
cut_off = cut_off
vocab = interp.vocab
_, tl_idx = interp.top_losses(len(interp.losses))
idxs = dl.get_losses(tl_idx, interp.preds, combs)
mc = interp.most_confused()
tbnames, boxes, cols, rows, ranges = self._get_names(x=dl,
idxs=idxs,
mc=mc,
varlist=varlist,
li=combs)
tb = widgets.TabBar(tbnames)
self._create_tabs(tb, tbnames, dl, interp, idxs, combs, boxes, cols,
rows, ranges, figsize, cut_off)
def plotting_region_consumption(_data_mapping, _rolling, _figsize):
_d = list(_data_mapping.keys())
tb = widgets.TabBar(_d, location='start')
for i in _d:
with tb.output_to(i):
_asu, _LOX, _LOX_ALSF, test_plot, test_stats = _data_mapping[i]
d_ = [
'enlèvements', 'covid 19', 'lits', 'réanimations',
'market share'
]
tb_ = widgets.TabBar(d_, location='top')
for j in d_:
with tb_.output_to(j):
if j == 'enlèvements':
try:
LOX_rolling_plot = _LOX[
'consommation totale'].rolling(_rolling).sum()
LOX_ALSF_rolling_plot = _LOX_ALSF[
'consommation totale'].rolling(_rolling).sum()
asu_rolling_plot = _asu.rolling(_rolling).sum()
asu_rolling_plot_ = list(asu_rolling_plot.columns)
asu_rolling_plot_ = [
'client LOX médical externes',
'client LOX médical internes'
] + asu_rolling_plot_
asu_rolling_plot[
'client LOX médical externes'] = LOX_rolling_plot
asu_rolling_plot[
'client LOX médical internes'] = LOX_ALSF_rolling_plot
asu_rolling_plot = asu_rolling_plot[
asu_rolling_plot_]
asu_rolling_plot[asu_rolling_plot.index.month ==
3].plot(figsize=_figsize)
except IndexError:
pass
elif j == 'covid 19':
test_plot.plot(figsize=_figsize)
else:
print(test_stats)
def displayTarBarImage(path_dir: Path,
first: int,
last: int,
start_last=False,
figsize: tuple = (3, 3)) -> None:
_enum_items = list(islice(itemize(path_dir), first, last))
tb_plt = widgets.TabBar([str(i) for i in range(first, last)],
location='start')
for base_0 in range(0, last - first):
with tb_plt.output_to(base_0, select=start_last):
print(f'File name: {Path(_enum_items[base_0][1].name)}')
print(
f'Image Size: {imageSize(Path(_enum_items[base_0][1]))} (H x W)'
)
displayImage(Path(_enum_items[base_0][1]), figsize=figsize)
def TabBar(*args):
"""Returns a real TabBar or a mock. Useful for UIs that don't support JS."""
if TABULATED_OUTPUT:
return widgets.TabBar(*args)
class MockTab:
def __init__(self):
pass
def __enter__(self):
pass
def __exit__(self, *x):
pass
class MockTabBar:
def __init__(self):
pass
def output_to(self, x):
return MockTab()
return MockTabBar()
def capa_vs_covid19_plotting(_data,
selection,
*,
figsize=(15, 15),
hspace=0.4,
wspace=0.1,
not_plotted=None):
if not_plotted is None:
not_plotted = list()
regions_departements_mapping = pd.read_csv(
'https://www.data.gouv.fr/en/datasets/r/987227fb-dcb2-429e-96af-8979f97c9c84'
)
regions_departements_mapping = regions_departements_mapping[[
'dep_name', 'region_name'
]]
gb_ = regions_departements_mapping.groupby('region_name')
dep_name_series = gb_['dep_name'].apply(list)
dep_name_series_mapping = dep_name_series.to_dict()
dep_name_series_mapping_plot = {
k: (v, max(math.ceil(len(v) / 2.0), 2))
for k, v in dep_name_series_mapping.items() if k not in not_plotted
}
_d = list(dep_name_series_mapping_plot.keys())
tb = widgets.TabBar(_d, location='start')
for i in _d:
with tb.output_to(i):
dep_set, rows = dep_name_series_mapping_plot[i]
try:
plot_capacity_vs_covid19(_data,
dep_set,
rows,
2,
selection,
figsize=figsize,
hspace=hspace,
wspace=wspace)
except IndexError:
pass
return True
def plotting_region_consumption_(_data_mapping, _data_dep, _rolling, _figsize):
_d = list(_data_mapping.keys())
tb = widgets.TabBar(_d, location='start')
for i in _d:
with tb.output_to(i):
_asu, _LOX, _LOX_ALSF, test_plot, test_stats, dep_set, map_lits, map_covid_19 = _data_mapping[
i]
d_ = [
'enlèvements', 'covid 19', 'lits', 'réanimations',
'market share'
]
tb_ = widgets.TabBar(d_, location='top')
for j in d_:
with tb_.output_to(j):
if j == 'enlèvements':
try:
LOX_rolling_plot = _LOX[
'consommation totale'].rolling(_rolling).sum()
LOX_ALSF_rolling_plot = _LOX_ALSF[
'consommation totale'].rolling(_rolling).sum()
asu_rolling_plot = _asu.rolling(_rolling).sum()
asu_rolling_plot_ = list(asu_rolling_plot.columns)
asu_rolling_plot_ = [
'client LOX médical externes',
'client LOX médical internes'
] + asu_rolling_plot_
asu_rolling_plot[
'client LOX médical externes'] = LOX_rolling_plot
asu_rolling_plot[
'client LOX médical internes'] = LOX_ALSF_rolling_plot
asu_rolling_plot = asu_rolling_plot[
asu_rolling_plot_]
asu_rolling_plot[asu_rolling_plot.index.month ==
3].plot(figsize=_figsize)
except IndexError:
pass
elif j == 'covid 19':
tb_dep = widgets.TabBar(dep_set, location='bottom')
data_plot_region = _data_dep[['capacité (2018)'
]].copy()
data_plot_region = data_plot_region[
data_plot_region.index.get_level_values(0).isin(
dep_set)]
data_plot_region = data_plot_region.groupby(
level=[1]).sum()
test_plot['capacité (2018)'] = data_plot_region[
'capacité (2018)']
test_plot = test_plot[[
'en reanimation ou soins intensifs',
'capacité (2018)', 'hospitalises', 'deces'
]]
for d in dep_set:
with tb_dep.output_to(d):
_fig, _axs = plt.subplots(nrows=1,
ncols=2,
figsize=_figsize)
plotting_figure(_axs[0], test_plot, title=i)
data_plot_ = _data_dep.loc[d, :][[
'reanimation', 'capacité (2018)'
]]
plotting_figure(_axs[1], data_plot_, title=d)
plt.show()
elif j == 'market share':
print(test_stats)
elif j == 'lits':
fig_dep_set, ax_dep_set = plt.subplots(1,
1,
figsize=(10,
10))
divider = make_axes_locatable(ax_dep_set)
cax = divider.append_axes("right", size="5%", pad=0.1)
fig_dep_set_ = map_lits.plot(column="lits",
cmap="Reds",
ax=ax_dep_set,
legend=True,
cax=cax)
for p in ['top', 'left', 'bottom', 'right']:
fig_dep_set_.axes.spines[p].set_visible(False)
fig_dep_set_.axes.xaxis.set_visible(False)
fig_dep_set_.axes.yaxis.set_visible(False)
elif j == 'réanimations':
fig_dep_set, ax_dep_set = plt.subplots(1,
1,
figsize=(10,
10))
divider = make_axes_locatable(ax_dep_set)
cax = divider.append_axes("right", size="5%", pad=0.1)
fig_dep_set_ = map_covid_19.plot(column="reanimation",
cmap="Reds",
ax=ax_dep_set,
legend=True,
cax=cax)
for p in ['top', 'left', 'bottom', 'right']:
fig_dep_set_.axes.spines[p].set_visible(False)
fig_dep_set_.axes.xaxis.set_visible(False)
fig_dep_set_.axes.yaxis.set_visible(False)
num_epochs=1,
shuffle=False))
predictions = []
for prediction_item, in zip(predictions_dict):
predictions.append(prediction_item['class_ids'][0])
actuals = list(test_df.loc[test_df[CATEGORY] == SUBGROUP]
['income_bracket'].apply(lambda x: '>50K' in x).astype(int))
classes = ['Over $50K', 'Less than $50K']
# To stay consistent, we have to flip the confusion
# matrix around on both axes because sklearn's confusion matrix module by
# default is rotated.
rotated_confusion_matrix = np.fliplr(confusion_matrix(actuals, predictions))
rotated_confusion_matrix = np.flipud(rotated_confusion_matrix)
tb = widgets.TabBar(['Confusion Matrix', 'Evaluation Metrics'], location='top')
with tb.output_to('Confusion Matrix'):
plot_confusion_matrix(rotated_confusion_matrix, classes)
with tb.output_to('Evaluation Metrics'):
grid = widgets.Grid(2, 3)
p, r, fpr = compute_eval_metrics(actuals, predictions)
with grid.output_to(0, 0):
print('Precision ')
with grid.output_to(1, 0):
print(' %.4f ' % p)
with grid.output_to(0, 1):
