[Python] Update column value of Pandas DataFrame

There are some ways to update column value of Pandas DataFrame.
So I wrote down them for the future reference.
I used documents of loc and where below as a reference.

Update column value of Pandas DataFrame

There are some ways to update column value of Pandas DataFrame.
Common ways are below.

• Bulk update by single value
• Update rows that match condition
• Update with another DataFrame

I'll introduce them with using DataFrame sample.

import pandas as pd

data_list1 = [
[1,2,3],
[2,3,4],
[3,4,5]
]
col_list1 = ["c1","c2","c3"]
df1 = pd.DataFrame(data=data_list1, columns=col_list1)
print(df1)

#    c1  c2  c3
# 0   1   2   3
# 1   2   3   4
# 2   3   4   5

Bulk update by single value

Bulk update is easy.
But it updates all rows. So it is not so useful and we can use it only for initialization.

df1["c3"]=1
print(df1)
#    c1  c2  c3
# 0   1   2   1
# 1   2   3   1
# 2   3   4   1

Update rows that match condition

If you want to update specific rows that match condition, you can use loc.

df1.loc[df1["c3"]>3, ["c3"]]=1
print(df1)
#    c1  c2  c3
# 0   1   2   3
# 1   2   3   1
# 2   3   4   1

Without using loc, you will see warning and you will not update values.

df1[df1["c3"]>3]["c3"]=1
print(df1)
# SettingWithCopyWarning: 
# A value is trying to be set on a copy of a slice from a DataFrame.
# Try using .loc[row_indexer,col_indexer] = value instead
#    c1  c2  c3
# 0   1   2   3
# 1   2   3   4
# 2   3   4   5

As an another method, we can update unmatched rows with using where.
In the example below, it updates column "c3" values of rows that have less values than 4 in column "c2".

df1["c3"].where(df1["c2"] >= 4, 10, inplace=True)  # Update unmatched cells
print(df1)

#    c1  c2    c3
# 0   1   2  10.0
# 1   2   3  10.0
# 2   3   4   1.0

Update with another DataFrame

There is a way to update column with another DataFrame.
With using this method, we can choose certail rows from parent DataFrame and apply updated values to parent DataFrame after the child process.

data_list2 = [
[5,7],
[6,8],
]
col_list2 = ["c4","c3"]
df2 = pd.DataFrame(data=data_list2, columns=col_list2)
print(df2)
#    c4  c3
# 0   5   7
# 1   6   8

df1.update(df2)
print(df1)
#    c1  c2   c3
# 0   1   2  7.0
# 1   2   3  8.0
# 2   3   4  1.0