Error in find pearson correlation function

[erneelgupta]

Output:
functional group  A_OS  A'_OS  A_HOMO-  ...      μĀ    μ𝐵 ̅       t  Perovskite
0              RbTaO3     1      1    0.000  ...  0.0000  0.0000  1.0815           1
1           Ba2PrIrO6     2      2    0.000  ...  0.0000  0.0804  0.9957           1
2           Ba2PrOsO6     2      2    0.000  ...  0.0000  0.0786  0.9946           1
3           Ba2SmIrO6     2      2    0.000  ...  0.0000  0.0196  1.0370           1
4           Ba2TbIrO6     2      2    0.000  ...  0.0000  0.0482  1.0172           1
...               ...   ...    ...      ...  ...     ...     ...     ...         ...
1500        Yb2NiMnO6     3      3    0.000  ...  0.0000  0.0571  0.9203           1
1501        Yb2NiRuO6     3      3    0.000  ...  0.0000  0.0250  0.9001           0
1502        GdBaCo2O6     3      2    1.067  ...  0.1193  0.0286  1.0205           1
1503        KLaMnWO6,     1      3    0.555  ...  0.0768  0.0089  1.0202           1
1504       NaTbMnWO6,     1      3   -0.264  ...  0.0439  0.0089  0.9492           1

[1505 rows x 36 columns]

The data set image is also attached.
This is the form of data set I have and I want to obtained Pearson's correlation of the file. I am running this in Jupyter Notebook.
The commands are as:

import numpy as np
import matplotlib.pyplot as plt     #.pyplot is a module
import pandas as pd
import seaborn as sns


x = pd.read_csv("formability_database.csv")

R1 = np.corrcoef(x, rowvar=False)

The error message is given below:

Error:
TypeError                                 Traceback (most recent call last)

<ipython-input-31-3c1804e8149c> in <module>()
----> 1 R1 = np.corrcoef(x, rowvar=False)
      2 #print(x.corr())

<__array_function__ internals> in corrcoef(*args, **kwargs)

3 frames

/usr/local/lib/python3.7/dist-packages/numpy/lib/function_base.py in corrcoef(x, y, rowvar, bias, ddof, dtype)
   2681         warnings.warn('bias and ddof have no effect and are deprecated',
   2682                       DeprecationWarning, stacklevel=3)
-> 2683     c = cov(x, y, rowvar, dtype=dtype)
   2684     try:
   2685         d = diag(c)

<__array_function__ internals> in cov(*args, **kwargs)

/usr/local/lib/python3.7/dist-packages/numpy/lib/function_base.py in cov(m, y, rowvar, bias, ddof, fweights, aweights, dtype)
   2516             w *= aweights
   2517 
-> 2518     avg, w_sum = average(X, axis=1, weights=w, returned=True)
   2519     w_sum = w_sum[0]
   2520 

<__array_function__ internals> in average(*args, **kwargs)

/usr/local/lib/python3.7/dist-packages/numpy/lib/function_base.py in average(a, axis, weights, returned)
    378 
    379     if weights is None:
--> 380         avg = a.mean(axis)
    381         scl = avg.dtype.type(a.size/avg.size)
    382     else:

/usr/local/lib/python3.7/dist-packages/numpy/core/_methods.py in _mean(a, axis, dtype, out, keepdims, where)
    180     if isinstance(ret, mu.ndarray):
    181         ret = um.true_divide(
--> 182                 ret, rcount, out=ret, casting='unsafe', subok=False)
    183         if is_float16_result and out is None:
    184             ret = arr.dtype.type(ret)

TypeError: unsupported operand type(s) for /: 'str' and 'int'

Larz60+ write Mar-01-2022, 04:27 PM:
Please post all code, output and errors (it it's entirety) between their respective tags. Refer to BBCode help topic on how to post. Use the "Preview Post" button to make sure the code is presented as you expect before hitting the "Post Reply/Thread" button.
Fixed for you this time. Please use BBCode on future posts.

[stevendaprano]

This is a good example of where the very last line of the error is sufficient to show the cause of the problem (although not necessarily to fix it!).

TypeError: unsupported operand type(s) for /: 'str' and 'int'

You are trying to work out the correlation coefficient of a column that includes strings.

numpy is trying to calculate the correlation coefficient of each pair of columns, but column 1 (functional groups) is all strings, so when numpy tries to compute the average, it adds up all those strings (concatenation) to get one large string, then tries to divide by the number of rows.

I'm not sure how to fix that, there doesn't seem to be any facility in np.corrcoef to select which columns (or rows) are used.

By the way, the code you have copied and pasted here seems to be different from the code in your screen shot.