how to read csv file and pass its values to a function parameters-CodePudding

I have a csv file having one column and the values it contains like this: here

and this function

def yolo_to_coco(x_center, y_center, w, h,  image_w, image_h):
  w = w * image_w
  h = h * image_h
  x1 = ((2 * x_center * image_w) - w)/2
  y1 = ((2 * y_center * image_h) - h)/2
  return [x1, y1, w, h]``

when I pass value to this function image_w and image_h I am passing from my own it is not in the csv

test = yolo_to_coco(53.854, 2912.798, 398.71, 57.202,1024,1024)

the get the required result

[-148993.02399999998, 2953417.7279999997, 408279.04, 58574.848]

I want to get each value from this file and pass to this function and the store in another csv file.

CodePudding user response：

You can use dataframe.apply(function) method. With this method you can apply this function to every row. Let's create a dummy df.

import pandas as pd
d = {'bbox': [[53.854, 2912.798, 398.71, 57.202],[31, 12,120, 20],[17, 72,170, 720]]}
df = pd.DataFrame(data=d)

But you don't need to create dummy df. You will read as

df = pd.read_csv ('/content/drive/MyDrive/bbox_file.csv')

let's create function. First argument will be row value. We will specify extra 2 arguments (image_w, image_h)

def yolo_to_coco(bbox,  image_w, image_h):
    row_list = list(bbox[1:-1].split(","))
    row_list = [float(i) for i in row_list]
    x_center, y_center, w, h = row_list
    w = w * image_w
    h = h * image_h
    x1 = ((2 * x_center * image_w) - w)/2
    y1 = ((2 * y_center * image_h) - h)/2
    return [x1, y1, w, h]

Now, We can apply this function to bbox column.You can pass any image_w,image_h values. It will create a pandas.series

df_coco_series = df["bbox"].apply(yolo_to_coco,image_w = 1024,image_h = 1024)

Finally converting to pandas.dataframe

df_coco = df_coco_series.to_frame()
print(df_coco)