-
Notifications
You must be signed in to change notification settings - Fork 2
/
visualizations.py
72 lines (48 loc) · 3.4 KB
/
visualizations.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
import seaborn as sns
import plotly.graph_objects as go
import plotly.express as px
import pandas as pd
import matplotlib.pyplot as plt
def plot_losses(train_loss_history, val_loss_history, num_epochs):
fig = go.Figure()
line_color = [px.colors.qualitative.Light24[5], px.colors.qualitative.Light24[4]]
fig.add_trace(go.Scatter(x=list(range(1, num_epochs+1)), y=train_loss_history, mode='lines', line=dict(color=line_color[0], width=1), name="train loss"))
fig.add_trace(go.Scatter(x=list(range(1, num_epochs+1)), y=val_loss_history, mode='lines',line=dict(color=line_color[1], width=1), name="valdiation loss")) #opacity=0.8
fig.update_yaxes(range=[0, 1])
fig.update_traces(textposition='top center')
fig.update_layout(autosize=False,width=900, height=500, title_text="SNN Loss", title_x=0.5, xaxis_title="Epoch", yaxis_title="", xaxis = dict(tickmode='linear', tick0=1, dtick=1), legend=dict(yanchor="top",xanchor="right", x=1.35, y=1),template="plotly_dark")
fig.show()
def plot_similarity_scores(non_matching_similarity, matching_similarity):
train_num_batchs = len(non_matching_similarity)
fig = go.Figure()
line_color = [px.colors.qualitative.Light24[22], px.colors.qualitative.Light24[19]]
fig.add_trace(go.Scatter(x=list(range(1, train_num_batchs+1)), y=non_matching_similarity, mode='lines', line=dict(color=line_color[0], width=1), name="unmatching categories"))
fig.add_trace(go.Scatter(x=list(range(1, train_num_batchs+1)), y=matching_similarity, mode='lines',line=dict(color=line_color[1], width=1), name="matching categories")) #opacity=0.8
fig.update_yaxes(range=[0, 1])
fig.update_traces(textposition='top center')
fig.update_layout(autosize=False,width=900, height=500, title_text="Similarity Scores", title_x=0.5, xaxis_title="Batch", yaxis_title="", legend=dict(yanchor="top",xanchor="right", x=1.35, y=1),template="plotly_dark")
fig.show()
def show_space(X, title, colors=None, color_by="", show_3D=False):
if show_3D:
dictionary = dict(zip(pd.DataFrame(X).columns, ["COMP1", "COMP2", "COMP3"]))
temp_df = pd.DataFrame(X).rename(columns=dictionary)
fig = px.scatter_3d(temp_df, x='COMP1', y='COMP2', z='COMP3',color=colors, template="plotly_dark", labels={"color": color_by})
else:
dictionary = dict(zip(X.columns, ["COMP1", "COMP2"]))
temp_df = pd.DataFrame(X).rename(columns=dictionary)
fig = px.scatter(temp_df, x='COMP1', y='COMP2',color=colors, template="plotly_dark", labels={"color": color_by})
fig.update_traces(marker=dict(size=4, opacity=0.98), textposition='top center')
fig.update_layout(title_text=title, title_x=0.5, autosize=False,width=900, height=500, legend=dict(yanchor="top",xanchor="right", x=1.1, y=1))
fig.show()
#fig.write_html("file.html")
def plot_confusion_matrix(mat,fig_size, labels):
fig = plt.figure(figsize=(fig_size,fig_size))
ax= fig.add_subplot(1,1,1)
sns.heatmap(mat, annot=True, cmap="Blues",ax = ax,fmt='g'); #annot=True to annotate cells
# labels, title and ticks
ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels');
ax.set_title('Confusion Matrix');
ax.xaxis.set_ticklabels(labels); ax.yaxis.set_ticklabels(labels);
plt.setp(ax.get_yticklabels(), rotation=30, horizontalalignment='right')
plt.setp(ax.get_xticklabels(), rotation=30, horizontalalignment='right')
plt.show()