Replacing append with concat for yolo predictions-CodePudding

I have read a bunch of posts related to the similar problems but I still don't understand how i can replace append with concat for my use case. I was using it for months as it is by just slapping this

warnings.simplefilter(action='ignore', category=FutureWarning)

car_model is a yolov5 model, whose results i have to iterate one by one, while confirming if the confidence is above a specific value. I just want to ask, does it even make sense for my use case to use concat instead of append since i have to iterate all the output ever single time anyway.

pred0 = car_model(frame, augment=False)
pred_df = pd.DataFrame()
for i, row in pred0.pandas().xyxy[0].iterrows():
   if row['confidence'] > confidence:
         pred_df = pred_df.append(row)

CodePudding user response：

You don't need to iterate over the dataframe. In fact, if the dataframe is huge then this would be very slow. You can avoid concatenating altogether by getting your pred_df like below:

pred0 = car_model(frame, augment=False)
pred0_df = pred0.pandas().xyxy[0]

pred_df = pred0_df[pred0_df['confidence'] > confidence]