Spaces:
Sleeping
Sleeping
add threshold on correlation values
Browse files
utils.py
CHANGED
@@ -11,13 +11,13 @@ def clean_csv_file(csv_file):
|
|
11 |
return df
|
12 |
|
13 |
|
14 |
-
def build_graph(csv_file):
|
15 |
features = clean_csv_file(csv_file)
|
16 |
links = features.corr(method='kendall').fillna(0).stack().reset_index()
|
17 |
links.columns = ['var1', 'var2', 'value']
|
18 |
|
19 |
# Keep only correlation over a threshold
|
20 |
-
links_filtered = links.loc[(links['value'] >
|
21 |
iter_values = iter(set(links_filtered['value']))
|
22 |
|
23 |
G = nx.from_pandas_edgelist(links_filtered, 'var1', 'var2')
|
@@ -46,9 +46,9 @@ class MplColorHelper:
|
|
46 |
return f"rgb({r},{g},{b})"
|
47 |
|
48 |
|
49 |
-
def display_graph(csv_file):
|
50 |
|
51 |
-
G = build_graph(csv_file)
|
52 |
|
53 |
CM_NAME = "Wistia"
|
54 |
|
|
|
11 |
return df
|
12 |
|
13 |
|
14 |
+
def build_graph(csv_file, threshold):
|
15 |
features = clean_csv_file(csv_file)
|
16 |
links = features.corr(method='kendall').fillna(0).stack().reset_index()
|
17 |
links.columns = ['var1', 'var2', 'value']
|
18 |
|
19 |
# Keep only correlation over a threshold
|
20 |
+
links_filtered = links.loc[(links['value'] > threshold) & (links['var1'] != links['var2'])]
|
21 |
iter_values = iter(set(links_filtered['value']))
|
22 |
|
23 |
G = nx.from_pandas_edgelist(links_filtered, 'var1', 'var2')
|
|
|
46 |
return f"rgb({r},{g},{b})"
|
47 |
|
48 |
|
49 |
+
def display_graph(csv_file, threshold):
|
50 |
|
51 |
+
G = build_graph(csv_file, threshold=threshold)
|
52 |
|
53 |
CM_NAME = "Wistia"
|
54 |
|