CODEWITHSUNDEEP
python-3.xpandasnumpy
tryingtobeastoic
tryingtobeastoic

Reputation: 195

How to remove all the values of a dataframe in the upper triangle including the diagonal values in pandas?

This is my dataframe:

                  id       age    gender    height    weight     ap_hi  \
id           1.000000  0.002623  0.003799  0.000221  0.000144  0.003489   
age          0.002623  1.000000 -0.018274 -0.077426  0.069705  0.018482   
gender       0.003799 -0.018274  1.000000  0.504722  0.130116  0.004941   
height       0.000221 -0.077426  0.504722  1.000000  0.248868  0.004300   
weight       0.000144  0.069705  0.130116  0.248868  1.000000  0.026527   
ap_hi        0.003489  0.018482  0.004941  0.004300  0.026527  1.000000   
ap_lo        0.000429  0.152787  0.059500  0.015356  0.223786  0.072260   
cholesterol  0.003867  0.129582 -0.037669 -0.064477  0.132686  0.022606   
gluc         0.002477  0.087280 -0.021178 -0.031410  0.104475  0.011004   
smoke       -0.002403 -0.044208  0.337682  0.187389  0.055805 -0.001978   
alco        -0.001039 -0.026956  0.169178  0.089257  0.058286  0.000607   
active       0.005890 -0.011471  0.007702 -0.005042 -0.012112 -0.000162   
cardio       0.003770  0.239987  0.001727 -0.025673  0.166886  0.050321   
overweight  -0.000769  0.089282 -0.055146 -0.156139  0.655764  0.016900   

                ap_lo  cholesterol      gluc     smoke      alco    active  \
id           0.000429     0.003867  0.002477 -0.002403 -0.001039  0.005890   
age          0.152787     0.129582  0.087280 -0.044208 -0.026956 -0.011471   
gender       0.059500    -0.037669 -0.021178  0.337682  0.169178  0.007702   
height       0.015356    -0.064477 -0.031410  0.187389  0.089257 -0.005042   
weight       0.223786     0.132686  0.104475  0.055805  0.058286 -0.012112   
ap_hi        0.072260     0.022606  0.011004 -0.001978  0.000607 -0.000162   
ap_lo        1.000000     0.148701  0.073920  0.022997  0.031839  0.002184   
cholesterol  0.148701     1.000000  0.383601  0.012797  0.037588  0.002804   
gluc         0.073920     0.383601  1.000000 -0.004203  0.013617 -0.009629   
smoke        0.022997     0.012797 -0.004203  1.000000  0.341434  0.027203   
alco         0.031839     0.037588  0.013617  0.341434  1.000000  0.026224   
active       0.002184     0.002804 -0.009629  0.027203  0.026224  1.000000   
cardio       0.326125     0.202257  0.088267 -0.020605 -0.011528 -0.037040   
overweight   0.169567     0.126770  0.086850 -0.003981  0.024210 -0.002382   

               cardio  overweight  
id           0.003770   -0.000769  
age          0.239987    0.089282  
gender       0.001727   -0.055146  
height      -0.025673   -0.156139  
weight       0.166886    0.655764  
ap_hi        0.050321    0.016900  
ap_lo        0.326125    0.169567  
cholesterol  0.202257    0.126770  
gluc         0.088267    0.086850  
smoke       -0.020605   -0.003981  
alco        -0.011528    0.024210  
active      -0.037040   -0.002382  
cardio       1.000000    0.141138  
overweight   0.141138    1.000000

This is a dataframe describing correlation, and I want to draw a heatmap. So, to remove redundancy, I want the upper triangular values including the diagonal values to be removed. What can I do?

Upvotes: 0

Views: 1160

Answers (1)

anon01
anon01

Reputation: 11171

You can use the numpy function to keep the lower triangular part of your matrix. Assuming it's square, this should work:

import numpy as np
df[:] = np.tril(df.values, k=-1)

input example:

              id       age    gender    height    weight     ap_hi

id      1.000000  0.002623  0.003799  0.000221  0.000144  0.003489
age     0.002623  1.000000 -0.018274 -0.077426  0.069705  0.018482
gender  0.003799 -0.018274  1.000000  0.504722  0.130116  0.004941
height  0.000221 -0.077426  0.504722  1.000000  0.248868  0.004300
weight  0.000144  0.069705  0.130116  0.248868  1.000000  0.026527
ap_hi   0.003489  0.018482  0.004941  0.004300  0.026527  1.000000

output:

              id       age    gender    height    weight  ap_hi

id      0.000000  0.000000  0.000000  0.000000  0.000000    0.0
age     0.002623  0.000000  0.000000  0.000000  0.000000    0.0
gender  0.003799 -0.018274  0.000000  0.000000  0.000000    0.0
height  0.000221 -0.077426  0.504722  0.000000  0.000000    0.0
weight  0.000144  0.069705  0.130116  0.248868  0.000000    0.0
ap_hi   0.003489  0.018482  0.004941  0.004300  0.026527    0.0

Upvotes: 1

Related Questions