如何使用索引列表从pandas数据帧中删除行

Question

Introduction

We have the following dataframe which we create from a CSV file.

data = pd.read_csv(path + name, usecols = ['QTS','DSTP','RSTP','DDATE','RDATE','DTIME','RTIME','DCXR','RCXR','FARE'] )

I want to delete specific rows from the dataframe. For this purpose I used a list and appended the ids of the rows we want to delete.

for index,row in data.iterrows():
     if (row['FARE'] >= 2500.00):
       indices.append(index)

From here i am lost. Don't know how to use the ids in the list to delete the rows from the dataframe

---

Question

• The list containing the row ids must be used in the dataframe to delete rows. Is it possible to do it?

---

Constraints

• We can't use data.drop(index,inplace=True) because it really slows the process
• We cannot use a filter because I have some special constraints.

Answer 1

If you are trying to remove rows that have 'FARE' values greater than or equal to zero, you can use a mask that have those values lesser than 2500 -

df_out = df.loc[df.FARE.values < 2500] # Or df[df.FARE.values < 2500]

For large datasets, we might want to work with underlying array data and then construct the output dataframe -

df_out = pd.DataFrame(df.values[df.FARE.values < 2500], columns=df.columns)

---

To use those indices generated from the loopy code in the question -

df_out = df.loc[np.setdiff1d(df.index, indices)]

Or with masking again -

df_out = df.loc[~df.index.isin(indices)]  # or df[~df.index.isin(indices)]

Answer 2

How about filtering data using DataFrame.query() method:

cols = ['QTS','DSTP','RSTP','DDATE','RDATE','DTIME','RTIME','DCXR','RCXR','FARE']
df = pd.read_csv(path + name, usecols=cols).query("FARE < 2500")