CODEWITHSUNDEEP
pythonpandas
ScalaBoy
ScalaBoy

Reputation: 3392

How to add multiple new columns with a fixed value?

I need to add columns to the DataFrame df. The values for all new columns should be fixed along all rows of df:

df = {
       "NUM":[1,2],
       "WAKE":["M","H"],
       "DISTANCE":[780,500]
     }

new_df = pd.DataFrame(df)

This is how I tried to add new multiple columns with fixed values.

for column, row in new_df.iterrows():
    row["TEMPERATURE"] = 20
    row["VISIBILITY"] = 5000
    row["WIND"] = 10

This code does not fail, but new columns are not created.

The expected result:

NUM   WAKE  DISTANCE  TEMPERATURE  VISIBILITY  WIND
1     M     780       20           5000        10
2     H     500       20           5000        10

Upvotes: 3

Views: 1136

Answers (4)

Ryan Ni
Ryan Ni

Reputation: 182

I had the same problem before trying to change values through iterrows. I realized that if you don't actually call your variable/data frame when making the change, you're not changing anything. I don't know the reasoning behind it, but for it to work you have to use .loc[column,column name] to make the actual change. 

for column, row in new_df.iterrows():
        new_df.loc[column,"TEMPERATURE"] = 20
        new_df.loc[column,"VISIBILITY"] = 5000
        new_df.loc[column,"WIND"] = 10

Upvotes: 1

ddrossi93
ddrossi93

Reputation: 385

Easy enough!

new_df['TEMPERATURE'] = 20
new_df['VISIBILITY'] = 5000
new_df['WIND'] = 10

and will yield:

  INDEX NUM WAKE  DISTANCE TEMPERATURE VISIBILITY WIND
    0    1    M     780       20         5000      10
    1    2    H     500       20         5000      10

This way you define the new columns for the dataframe and set each row value to the single value

Upvotes: 2

BENY
BENY

Reputation: 323316

Two steps 

adddf=pd.DataFrame({'TEMPERATURE': 20, 'VISIBILITY': 5000, 'WIND': 10},index=new_df.index)
pd.concat([new_df,adddf],1)
Out[253]: 
   DISTANCE  NUM WAKE  TEMPERATURE  VISIBILITY  WIND
0       780    1    M           20        5000    10
1       500    2    H           20        5000    10

Upvotes: 2

cs95
cs95

Reputation: 402813

This is as simple as a single assign call with a dictionary:

dct = {'TEMPERATURE': 20, 'VISIBILITY': 5000, 'WIND': 10}
new_df2 = new_df.assign(**dct)
new_df2

   NUM WAKE  DISTANCE  TEMPERATURE  VISIBILITY  WIND
0    1    M       780           20        5000    10
1    2    H       500           20        5000    10

Upvotes: 5

Related Questions