CODEWITHSUNDEEP
pythonpandasnumpydataframefor-loop
Ngan Huynh
Ngan Huynh

Reputation: 5

How to store generate result from a function inside a for loop to DataFrame in Python?

I would like to store generate results from a function that I created called "surface_coverage". The function returns result in numpy.ndarray form and I tried to store it into a dataframe but I kept getting error msg "Must pass 2-d input" here is my code 

T = [400,500,600,700]
result = []
for t in  df_G['Temperature']:
    for i in T:

        columns = f'G_ads_{i}'
        G_CO2 = df.loc[df.Metal_Oxides == "Al2O3"].loc[df.Adsorbates == 'CO2'][columns].min()
        G_H2O = df.loc[df.Metal_Oxides == "Al2O3"].loc[df.Adsorbates == 'H2O'][columns].min()
        G_O2 = df.loc[df.Metal_Oxides == "Al2O3"].loc[df.Adsorbates == 'O2'][columns].min()
        if t == i+273.15:
            theta = surface_coverage(t,P,G_CO2,G_H2O,G_O2,x_co2,x_h2o,x_o2)
            result.append(theta)
new_data = pd.DataFrame(result)

and here is the output:

[array([[8.53931326e-04],
       [9.34890812e-15],
       [9.99146054e-01],
       [1.46447007e-08]]), array([[1.07403011e-01],
       [4.44545478e-13],
       [8.92596825e-01],
       [1.64041799e-07]]), array([[8.52759436e-01],
       [1.52248154e-12],
       [1.47240375e-01],
       [1.88472871e-07]]), array([[9.92480337e-01],
       [8.43223552e-13],
       [7.51961814e-03],
       [4.46422474e-08]]), array([[9.99428328e-01],
       [4.36060531e-13],
       [5.71659951e-04],
       [1.17418613e-08]]), array([[9.99935140e-01],
       [2.38117323e-13],
       [6.48560836e-05],
       [3.70506323e-09]])]

Is there any method to convert numpy.bdarray into float? Thank you for your time and your help!

Upvotes: 0

Views: 59

Answers (1)

ryl
ryl

Reputation: 86

I think the issue here is that your output array is nested in a way that looks 3D rather than 2D, which is what the pandas DataFrame is looking for.

It looks like theta, the output of your surface_coverage function, is a matrix. While you might be interpreting it is as a list of 4 numbers (a vector), it looks like it is being treated as a matrix (1 row x 4 columns, which we can see by counting the brackets). 

array([[8.53931326e-04],
   [9.34890812e-15],
   [9.99146054e-01],
   [1.46447007e-08]])

You probably want this array to look like this (note the brackets):

array([8.53931326e-04,
   9.34890812e-15,
   9.99146054e-01,
   1.46447007e-08])

To fix this, you should be able to replace 

result.append(theta)

with

result.append(theta.flatten())

That'll turn these into just arrays. Your result will therefore just be a list of arrays which it should correctly identify as 2D.

Upvotes: 1

Related Questions