def plot(masking: "Masking", mplot: plt) -> plt:
"""
Plot layer wise density bar plot.
:param masking: Masking instance
:type masking: sparselearning.core.Masking
:param mplot: matplotlib object
:type mplot: pyplot
:return: matplotlib plot
:rtype: pyplot
"""
density_ll = _get_density_ll(masking)
bin_ll = np.arange(len(density_ll)) + 1
width = 0.8
mplot.clf()
mplot.bar(bin_ll, density_ll, width, color="b")
# Gets too crowded when including layer names
# layer_name_ll = list(masking.masks.keys())
# plt.xticks(bin_ll, layer_name_ll)
mplot.ylabel("Density")
mplot.xlabel("Layer Number")
return mplot
def generate(self, plot: pyplot, data: Any) -> None:
# [("gcc", "ref", 184585549)]
series: Dict[str, List[float]] = {}
for row in data:
label = " ".join(row[0:2])
if label not in series:
series[label] = []
series[label].append(row[2])
values = list(series.values())
values.sort(key=lambda x: numpy.std(x), reverse=True)
keys = list(series.keys())
keys.sort(key=lambda x: numpy.std(series[x]), reverse=True)
plot.boxplot(values, showfliers=False)
plot.gcf().axes[0].yaxis.get_major_formatter().set_scientific(False)
plot.gcf().axes[0].set_xticklabels(keys)
plot.title("")
plot.ylabel(self.options.event)
plot.xlabel("Optimizations")
def generate(self, plot: pyplot, data: Any) -> None:
# [("Modern Workstation", "gcc", "ref", 34579902)]
series: Dict[str, List[float]] = {}
for row in data:
label = "{} {}".format(row[1], row[2])
if label not in series:
series[label] = []
series[label].append(row[3] / 1e6)
values = list(series.values())
values.sort(key=lambda x: min(x), reverse=True)
keys = list(series.keys())
keys.sort(key=lambda x: min(series[x]), reverse=True)
plot.boxplot(values, showfliers=False)
plot.gcf().axes[0].yaxis.get_major_formatter().set_scientific(False)
plot.gcf().axes[0].set_xticklabels(keys)
plot.title("")
plot.ylabel("Duration (ms)")
plot.xlabel("Optimizations")
def plot3D(ax: plt, sub3d: plt, X: np.ndarray, y: np.ndarray, w: np.ndarray,
name: str) -> None:
'''
Visualize decision boundary and data classes in 3D
:param ax: matplotlib
:param sub3d: fig.add_subplot(XXX, projection='3d')
:param X: data
:param y: data labels
:param w: model parameters
:param name: plot name identifier
:return:
'''
x1 = np.array(X[1, :]) # note: X_train[0,:] is the added row of 1s (bias)
x2 = np.array(X[2, :])
posterior1 = LOGREG().activationFunction(w, X)
posterior1 = np.squeeze(np.asarray(posterior1))
markers = ['o', '+']
groundTruthLabels = np.unique(y)
for li in range(len(groundTruthLabels)):
x1_sub = x1[y[:] == groundTruthLabels[li]]
x2_sub = x2[y[:] == groundTruthLabels[li]]
m_sub = markers[li]
posterior1_sub = posterior1[y[:] == groundTruthLabels[li]]
sub3d.scatter(x1_sub,
x2_sub,
posterior1_sub,
c=posterior1_sub,
vmin=0,
vmax=1,
marker=m_sub,
label='ground truth label = ' + str(li))
ax.legend()
x = np.arange(x1.min(), x1.max(), 0.1)
pms = [[0.1, 'k:'], [0.25, 'k--'], [0.5, 'r'], [0.75, 'k-.'], [0.9, 'k-']]
for (p, m) in pms:
yp = (-np.log((1 / p) - 1) - w[1] * x - w[0]) / w[2]
yp = np.squeeze(np.asarray(yp))
z = np.ones(yp.shape) * p
sub3d.plot(x, yp, z, m, label='p = ' + str(p))
ax.legend()
ax.xlabel('feature 1')
ax.ylabel('feature 2')
ax.title(name + '\n Posterior for class labeled 1')
def plot_choice(num: int, plot: plt, df1: pd.DataFrame) -> None:
"""
Function to re-draw the canvas
:param num: int of the RadioButton selection
:param plot: plt Matplotlib.pyplot instance
:param df1: pd.DataFrame of the TSA data
:return: None
"""
plot.clf()
blue = "#327ff6"
gray = "#bdb8b6"
if num == 1:
# Daily raw pax numbers
fig, axis = plt.subplots()
axis.plot(df1['Date'], df1['2020'], color=blue, linestyle='-', label="2020")
axis.plot(df1['Date'], df1['2019'], color=gray, linestyle='-', label="2019")
plot.legend()
plot.xlabel('Date')
plot.ylabel('Passengers')
plot.title('TSA Passenger Daily Throughput 2019 and 2020')
axis.get_yaxis().set_major_formatter(FuncFormatter(lambda x, p: format(int(x), ',')))
plot_new_canvas(fig, root)
elif num == 2:
# Weekly raw pax numbers
weekly_data = df1.resample('W-Mon', label='right', closed='right',
on='Date').sum().reset_index().sort_values(by='Date')
fig, axis = plt.subplots()
axis.plot(weekly_data['Date'], weekly_data['2020'], color=blue, linestyle='-',
label="2020")
axis.plot(weekly_data['Date'], weekly_data['2019'], color=gray, linestyle='-',
label="2019")
axis.get_yaxis().set_major_formatter(FuncFormatter(lambda x, p: format(int(x), ',')))
plot.xlabel('Date')
plot.ylabel('Passengers')
plot.legend()
plot.ylim(0, 20E6)
plot.title('TSA Passenger Weekly Throughput 2019 and 2020')
plot_new_canvas(fig, root)
elif num == 3:
# Daily YoY percentage
fig, axis = plt.subplots()
axis.plot(df1['Date'], df1['yoy'], color=blue, linestyle='-', label="2020")
plot.legend()
plot.xlabel('Date')
plot.ylabel('Passenger Load Factor (%)')
plot.title('TSA Passenger Daily Throughput in 2020 as a Percentage of 2019')
plot_new_canvas(fig, root)
elif num == 4:
# Weekly YoY percentage
weekly_data = df1.resample('W-Mon', label='right', closed='right',
on='Date').sum().reset_index().sort_values(by='Date')
weekly_data['yoy'] = (weekly_data['2020']/weekly_data['2019'])*100
fig, axis = plt.subplots()
axis.plot(weekly_data['Date'], weekly_data['yoy'], color=blue, linestyle='-',
label="2020")
plot.xlabel('Date')
plot.ylabel('Passenger Load Factor (%)')
plot.legend()
plot.title('TSA Passenger Daily Throughput in 2020 as a Percentage of 2019')
plot_new_canvas(fig, root)
def generate(self, plot: pyplot, data: Any) -> None:
# [('low-end-laptop', 0, 'mceliece', '6960119f', 'clang', 'ref-optimized', 'keypair', 666.1022, 999, 665165462)]
# clang, ref-optimized: [1, 2, 3, ...]
series: Dict[str, Dict[int, float]] = {}
# run_index: average duration
baseline_average_durations: Dict[int, int] = {}
# label: run_index: iteration: duration
sum_per_label: Dict[str, Dict[int, Dict[int, int]]] = {}
# Find values
for row in data:
compiler = row[4]
features = row[5]
run_index = row[1]
if compiler == "gcc" and features == "ref":
average_duration = row[7]
baseline_average_durations[run_index] = average_duration
else:
label = "{} {}".format(compiler, features)
if label not in sum_per_label:
sum_per_label[label] = {}
if run_index not in sum_per_label[label]:
sum_per_label[label][run_index] = {}
iteration = row[8]
duration = row[9]
sum_per_label[label][run_index][iteration] = duration
baseline_avarage_duration = sum(baseline_average_durations.values()
) / len(baseline_average_durations)
print("Baseline runs:", baseline_average_durations)
print("Baseline average duration:", baseline_avarage_duration)
for label in sum_per_label.keys():
max_length = max([
len(value.items()) for value in sum_per_label[label].values()
])
for i in range(max_length):
relevant_items = [
items[i] for items in sum_per_label[label].values()
if i in items
]
duration = sum(relevant_items) / len(relevant_items)
percentual_duration = (duration /
1e6) / baseline_avarage_duration
if label not in series:
series[label] = {}
series[label][i] = (1 / percentual_duration - 1.0)
colors = [
"#e6194B", "#3cb44b", "#4363d8", "#f58231", "#800000", "#9A6324",
"#000075", "#469990"
]
for i, key in enumerate(series.keys()):
# TODO: may be wrong if there are gaps in data as it does not care about the acutal indexing
values = series[key].values()
plot.plot(values, label=key, color=colors[i])
plot.title("")
plot.ylabel("Speedup")
plot.xlabel("Iteration")
plot.legend(bbox_to_anchor=(0.5, 1.05),
loc="lower center",
fontsize=8,
ncol=len(series))
def _set_x_label(cls, p: plt, label: str):
p.xlabel(label, fontsize=36)
