How to check linear reliance between dependent variable and independent variables? Because so as to make Linear Regression model in Python we have to use (as I suppose) only variables which are: 1. correlated with dependent variable 2. independent variables which are not correlated with other independent variables 3. independent variables with linear reliance with dependent variables ? Please give me the code which is able to chech linear reliance in Python
Asked
Active
Viewed 107 times
-1
-
Your question is not very clear, but if you use Pandas you can try function corr() to check correlations between all valiables in dataset: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.corr.html – CrazyElf Dec 27 '19 at 10:13
-
You can also check this topic: https://stackoverflow.com/questions/29432629/plot-correlation-matrix-using-pandas – CrazyElf Dec 27 '19 at 10:15
1 Answers
0
You can use pandas for that:
df = pd.DataFrame({'feature one': [1,2,3,5,2,3],
'feature two': [5,10,18,23,16,20],
'feature three': [-23,-4,1,29,2,112],
'result': [10,20,30,50,20,30]})
print(df)
print(df.corr())
You will see that feature one
has the biggest correlation with result, then feature two
and then feature three
. You can also check out the correlation between each feature.
So, for your linear model I would chose first and second feature
If values are close to -1 or 1, that means that there is big correlation between features, if values are close to 0, that means that that there is no correlation.

taga
- 3,537
- 13
- 53
- 119
-
taga, so as I mean if values are close to 0 the is NOT a linear reliance between independent variables and dependent variable? – dingaro Dec 27 '19 at 10:39
-
If values in correlation result are close to 0 , that means that there is no correlation (use classification algorithms). – taga Dec 27 '19 at 10:40
-
-