CODEWITHSUNDEEP
pythonpandasreplace
Rita
Rita

Reputation: 13

How to set or replace values of rows by a previous column value if conditions exist in python pandas

I'm trying to solve an issue related to set or replace values in rows by the value of the previous one in certain conditions. The data looks like:

Point_ID [1,2,3,1,2,1], Shape_ID [84,85,86,87,88,89]

I used mask and shift in conditions if Point_ID starts with 1 the Shape_ID of the second row should be replaced by the value of the previous row in case the number increases. If not then Shape_ID should stay in a row. So I am trying to obtain:

Point_ID[1,2,3,1,2,1], Shape_ID[84,84,84,87,87,89]

Upvotes: 1

Views: 60

Answers (2)

BENY
BENY

Reputation: 323226

Using eq,cumsum

df['GroupId']=df.Point_ID.eq(1).cumsum()
df.Shape_ID[df.Point_ID.ne(1)]=np.nan
df.assign(Shape_ID=df.groupby('GroupId')['Shape_ID'].ffill()).drop('GroupId',1)

Out[831]: 
   Point_ID  Shape_ID
0         1      84.0
1         2      84.0
2         3      84.0
3         1      87.0
4         2      87.0
5         1      89.0

Upvotes: 0

Scott Boston
Scott Boston

Reputation: 153460

Use mask, ne, and ffill:

df= pd.DataFrame({'Point_ID':[1,2,3,1,2,1] , 'Shape_ID': [84,85,86,87,88,89]})
print(df)

   Point_ID  Shape_ID
0         1        84
1         2        85
2         3        86
3         1        87
4         2        88
5         1        89

df.assign(Shape_ID=df['Shape_ID'].mask(df['Point_ID'].ne(1)).ffill().astype(int))

Output:

   Point_ID  Shape_ID
0         1        84
1         2        84
2         3        84
3         1        87
4         2        87
5         1        89

Upvotes: 2

Related Questions