CODEWITHSUNDEEP
pythonpandasdataframe
WangYang.Cao
WangYang.Cao

Reputation: 143

Change column value on rows which fulfilled conditions in DataFrame

Guys,

I have a df like this :

A       B       C        D       E      Yes
14      12     123     153      178       0
13      1      435      55       87       0
14      12     123       1      435       0
......

15      0      125      66       90       0

Let us say, we have two variable x and y, are integer. I want to change the 'Yes' column to '1' if any one of following conditions are fulfilled :

    df.D < x and df.E > x
    df.D > x and df.E > y
    df.D > y and df.E > y

Besides, I am sure df.E is always larger than df.D in those raw data.

How can I do it quickly ? I tried to write some expression based on that, but all have some problems ... Really appreciate.

Upvotes: 1

Views: 48

Answers (2)

Sociopath
Sociopath

Reputation: 13401

Addition to @jpp answer you can use np.wheretoo

df = pd.DataFrame({'A':[14,13,14,15], 'B':[12,1,12,0], 'D':[153,55,1,66],'E':[178,87,435,90],'Yes':[0,0,0,0]})

x = 100
y = 200

m1 = (df['D'] < x) & (df['E'] > x)
m2 = (df['D'] > x) & (df['E'] > y)
m3 = (df['D'] > y) & (df['E'] > y)


df['Yes'] = np.where(m1|m2|m3, 1, 0)

print(df)

Output:

    A   B    D    E  Yes
0  14  12  153  178    0
1  13   1   55   87    0
2  14  12    1  435    1
3  15   0   66   90    0

Upvotes: 0

jpp
jpp

Reputation: 164623

You can create some Boolean series and use them as masks with pd.DataFrame.loc. For example:

x = 10
y = 20

m1 = (df['D'] < x) & (df['E'] > x)
m2 = (df['D'] > x) & (df['E'] > y)
m3 = (df['D'] > y) & (df['E'] > y)

df.loc[m1 | m2 | m3, 'Yes'] = 1

Upvotes: 1

Related Questions