tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

adde dummy classifier

Browse source
This commit is contained in:
Tanushree Tunstall 2022-07-07 12:28:58 +01:00
parent a15d801c2a
commit d14c3f9c4a
1 changed files with 97 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

97
scripts/ml/dummy_classifier.py Normal file
View file

@ -0,0 +1,97 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu Jul 7 10:54:09 2022
@author: tanu
"""
import numpy as np
from sklearn.dummy import DummyClassifier
X_eg = np.array([-1, 1, 1, 1])
y_eg = np.array([0, 1, 1, 1])
dummy_clf = DummyClassifier(strategy="most_frequent")
dummy_clf.fit(X_eg, y_eg)
DummyClassifier(strategy='most_frequent')
dummy_clf.predict(X_eg)
np.array([1, 1, 1, 1])
dummy_clf.score(X_eg, y_eg)
0.75
dummy_clf.matthews_corrcoef(X_eg, y_eg)
#%%
df['dst_mode']
y_all_tt = df.loc[:,'dst']
y_all_tt.value_counts()
#Counter(y_all_tt)
#0: 71, 1: 114
y_all_tt.value_counts(normalize = True)
df2['y']
y_train_tt = df2['y']
Counter(y_train_tt)
##0: 41, 1: 82
y_train_tt.value_counts(normalize = True)
df2['y_bts']
y_bts_tt = df2['y_bts']
Counter(y_bts_tt)
#0: 21, 1: 41
y_bts_tt.value_counts(normalize = True)
#%%
df_clean = df[df['dst'].notna()]
X = df_clean.iloc[:,0:171]
X.columns
y = df_clean.iloc[:,171] # dst
y.value_counts()
y2 = df_clean.iloc[:,172] #dst_mode
y2.value_counts()
X_train_tt,X_test_tt, y_train_tt, y_test_tt = train_test_split(X, y, test_size=0.30, random_state=42, stratify = y)
y2.value_counts()
round(y.value_counts(normalize = True),2)
y_train_tt.value_counts()
round(y_train_tt.value_counts(normalize = True),2)
y_test_tt.value_counts()
round(y_test_tt.value_counts(normalize = True),2)
dummy_clf = DummyClassifier(strategy="most_frequent")
dummy_clf.fit(X_train_tt, y_train_tt)
DummyClassifier(strategy='most_frequent')
dummy_clf.predict(X_test_tt)
# pnca: split 0/30
=======================
Total y count in data:
1.0 114 (62%)
0.0 71 (38%)
=======================
=======================
Train y count in data:
1.0 79 (61%)
0.0 50 (39%)
=======================
=======================
Test y count in data:
1.0 35 (62%)
0.0 21 (38%)
=======================
acccuracy:
TP+TN/TP+TN+FP+FN
114/71