-
Notifications
You must be signed in to change notification settings - Fork 0
/
app.py
205 lines (185 loc) · 5.9 KB
/
app.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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
import polars as pl
from plotly import express as px
import yaml
import streamlit as st
from dash_bio import Clustergram
from sklearn.preprocessing import RobustScaler
from sklearn.pipeline import Pipeline
import umap
def main():
title = read_config_value("text", "title")
st.set_page_config(
page_title=title,
page_icon=":ballot_box_with_ballot:",
menu_items={
"Report a bug": "https://github.com/Excidion/wahl-o-mat/issues/new",
"About": """
+ [Author](https://www.linkedin.com/in/cedricwilting/) on LinkedIn
+ [Source Code](https://github.com/Excidion/wahl-o-mat) on GitHub
""",
},
)
st.title(title)
st.write(read_config_value("text", "intro"))
election = st.selectbox("Wahl", get_dataset_names())
df = load_dataset(election)
st.write(read_config_value("text", "heatmap0"))
st.plotly_chart(
plot_heatmap(df), use_container_width=True, config={"displayModeBar": False}
)
st.write(read_config_value("text", "heatmap1"))
st.subheader("Parteienlandkarte")
st.write(read_config_value("text", "clusters0"))
plot_party_clusters(df)
st.write(read_config_value("text", "clusters1"))
def plot_heatmap(df):
matrix = df.pivot(
index="topic",
columns="party",
values="opinion",
aggregate_function=None, # raise error on duplicates
)
matrix = matrix.to_pandas()
y = matrix.pop("topic")
# remove all neutral parties and topics
matrix = matrix.loc[:, ~(matrix == 0).all(axis=0)]
matrix = matrix.loc[~(matrix == 0).all(axis=1)]
# plot
fig = Clustergram(
data=matrix,
column_labels=list(matrix.columns),
row_labels=list(y),
# make dedograms disappear
display_ratio=0,
line_width=0,
# color the heatmap
color_map=[
[0, "orange"], # "rgb(217, 95, 2)"],
[0.5, "black"], # "rgb(102, 102, 102)"],
[1, "blue"], # "rgb(27, 158, 119)"],
],
center_values=False, # show real data
row_dist="cosine",
col_dist="cosine",
link_method="ward",
)
# remove colorbar
fig.data[-1].update(showscale=False)
return fig
def plot_party_clusters(df):
config = {
"selectZoom": False,
"scrollZoom": True,
"displayModeBar": False,
}
dims = st.radio(
"Dimension der Karte", [3, 2], horizontal=True, format_func=lambda x: f"{x}D"
)
st.plotly_chart(
_plot_party_clusters(df, dims), use_container_width=True, config=config
)
# @st.cache_data
def _plot_party_clusters(_df, dimensions=2):
opinions = _df.pivot(
index="party",
columns="topic",
values="opinion",
aggregate_function=None, # raise error on duplicates
)
opinions = opinions.to_pandas()
parties = opinions.pop("party")
pipe = Pipeline(
[
("scaler", RobustScaler()),
(
"umap",
umap.UMAP(
n_components=dimensions,
n_neighbors=3,
random_state=42,
metric="cosine",
),
),
]
)
embedding = pipe.fit_transform(X=opinions)
opinions["x"] = embedding[:, 0]
opinions["y"] = embedding[:, 1]
if dimensions == 3:
opinions["z"] = embedding[:, 2]
opinions["party"] = parties
color_discrete_map = {
"CDU / CSU": "#151518",
"PIRATEN": "#ff820a",
"GRÜNE": "#409A3C",
"AfD": "#009ee0",
"SPD": "#E3000F",
"FDP": "#FFED00",
"DIE LINKE": "#BE3075",
"Volt": "#562883",
"Die PARTEI": "#b5152b",
"BÜNDNIS DEUTSCHLAND": "#a2bbf3",
"FAMILIE": "#ff6600",
"FREIE WÄHLER": "#F7A800",
}
for party in parties:
if party not in color_discrete_map:
color_discrete_map[party] = "#696969"
plot_args = dict(
hover_name="party",
labels={
"x": "",
"y": "",
"z": "",
},
color="party",
color_discrete_map=color_discrete_map,
)
if dimensions == 2:
fig = px.scatter(opinions, x="x", y="y", **plot_args)
elif dimensions == 3:
plot_args["labels"]["z"] = ""
fig = px.scatter_3d(opinions, x="x", y="y", z="z", **plot_args)
else:
raise ValueError("Parameter 'dimensions' has to be either 2 or 3.")
fig.update_xaxes(tickvals=[], zeroline=False)
fig.update_yaxes(tickvals=[], zeroline=False)
fig.update_layout(showlegend=False)
fig.update_traces(marker=dict(line=dict(width=0.5, color="white")))
return fig
def get_dataset_names():
datasets = read_config().get("datasets").keys()
return [d for d in datasets if d != "_defaults"]
def load_dataset(name):
settings = load_settings(name)
df = pl.read_excel(settings.get("file"), **settings.get("kwargs"))
for k, v in settings.get("columns").items():
df = df.rename({v: k})
for col, mapping in settings.get("mapping").items():
default = mapping.pop("_default", None)
df = df.with_columns(pl.col(col).replace(mapping, default=default))
# chop too long party names
df = df.with_columns(
pl.when(pl.col("party").str.len_chars() > 20)
.then(pl.col("party").str.split(by=" ").list.last())
.otherwise(pl.col("party"))
.alias("party")
)
return df
def load_settings(name):
config = read_config()
datasets = config.get("datasets")
settings = datasets.get("_defaults")
dataset = datasets.get(name)
settings.update(dataset)
return settings
def read_config_value(*args):
config = read_config()
for key in args:
config = config.get(key)
return config
def read_config():
with open("config.yml", "r") as infile:
return yaml.safe_load(infile)
if __name__ == "__main__":
main()