Store np.float64 and np.array values as column values in dataframe

Question

I have two numpy arrays and dataframe as given below

val = np.array([0.501,0.32])
values = np.arange(24).reshape((2,3,4))
input_df = pd.DataFrame(columns=['colname_' + str(i) for i in range(4)])

I would like to

a) Create a new dataframe (dummy) with 3 columns such as ROW_ID, FEATURE NAME, Contribution

b) values for dummy dataframe should be populated using np.array above and column names from input_df`

c) Under the Feature Name column use the input_df column names

b) Populate the val[0] as contribution in dummy dataframe and also use each element from values[0][1] to populate it in contribution column. I tried the below code

pd.DataFrame({
        "Feature Name": ["Base value"] + [f"{col}" for col in df.columns.tolist()],
        "Contribution": (val[0].tolist()) + list(values[0][1])
    })

But I get an error message

TypeError: unsupported operand type(s) for +: 'float' and 'list'

Or I also receive another error which is

ValueError: All arrays must be of the same length

I expect my output to be like as shown below

update - real data issue

Panwen Wang · Accepted Answer · 2022-03-31 16:31:39Z

1

Try:

pd.DataFrame({
  "Feature Name": ["Base value"] + [f"{col}" for col in df.columns.tolist()],
  "Contribution": (val[:1].tolist()) + list(values[0][1])
  #                   ^^^^
})

val[0] makes it a scalar value, even followed by .tolist()

>>> type(val[0].tolist())
<class 'float'>

answered Mar 31, 2022 at 16:31

Panwen Wang

3,8652 gold badges21 silver badges41 bronze badges

Sign up to request clarification or add additional context in comments.

7 Comments

The Great Over a year ago

thanks, upvoted for the help. what does val[:1] mean? It takes all values upto index 1?

Panwen Wang Over a year ago

Correct. It slices the list with the first element. You could also do [val[0]]

Panwen Wang Over a year ago

You don't need that extra [] around (explainer.expected_value[:1].tolist()), it's already been a list. Same for the latter part shape_values[0][1].

The Great Over a year ago

I missed a list function. thanks for your help. It worked

The Great Over a year ago

Do you know how can I do this for multiple rows in dataframe of 2000 rows? Currently, we did for one row only (ROW_ID = 20). We need to do similarly for 1999 rows. Can you help me with it? I will open a new question?

|

Collectives™ on Stack Overflow

Store np.float64 and np.array values as column values in dataframe

1 Answer 1

7 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

7 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related