Since you can't insert the row and push others back directly, a clever trick you can use is create a new order:
# adds a new column, "new" with the original order
df['new'] = range(1, len(df) + 1)
# sets value that has index 53 with 0 on the new column
# note that this comparison requires you to match index type
# so if weeks are object, you should compare df.index == '53'
df.loc[df.index == 53, 'new'] = 0
# sorts values by the new column and drops it
df = df.sort_values("new").drop('new', axis=1)
Before:
numbers
weeks
1 181519.23
2 18507.58
3 11342.63
4 6064.06
53 4597.90
After:
numbers
weeks
53 4597.90
1 181519.23
2 18507.58
3 11342.63
4 6064.06