如何拆分LINESTRING坐标数据并生成指定格式的Pandas DataFrame？

Absolutely! Let's break down how to solve both of your requirements step by step:

Step 1: Fix the Initial Data Format #

First, note that your original DataFrame example uses a raw LINESTRING keyword which isn't valid Python syntax—we'll assume your actual col1 values are string representations of the geometry (like "LINESTRING(...)"). Here's the corrected starting point:

import pandas as pd
import re

# Corrected input DataFrame (col1 contains string values)
d = {'col1': ["LINESTRING(174.76028 -36.80417,174.76041 -36.80389, 175.76232 -36.82345)"]}
df = pd.DataFrame(d)

Step 2: Define Your Custom Value-Processing Function #

You can tailor this function to any logic you need (rounding, scaling, transformations, etc.). For example, let's use a function that rounds coordinates to 2 decimal places:

# Custom function to process individual coordinate values
def process_coord_value(val):
    # Example: Round to 2 decimal places; replace with your logic here
    return round(float(val), 2)

Step 3: Parse LINESTRING and Generate Target Columns #

We'll create a helper function to extract the geometry type (like LINESTRING) and process each coordinate pair, then apply it to your DataFrame to split into col1 and col2:

def parse_linestring(geom_str):
    # Extract the geometry type (e.g., LINESTRING)
    geom_type = re.match(r'^(\w+)', geom_str).group(1)
    # Extract all coordinate pairs inside the parentheses
    coords_raw = re.findall(r'\((.*?)\)', geom_str)[0].split(',')
    
    # Process each coordinate pair and convert to tuples
    processed_coords = []
    for pair in coords_raw:
        x_str, y_str = pair.strip().split()
        processed_x = process_coord_value(x_str)
        processed_y = process_coord_value(y_str)
        processed_coords.append((processed_x, processed_y))
    
    return geom_type, processed_coords

# Apply the parser to col1 and expand into two new columns
df[['col1', 'col2']] = df['col1'].apply(lambda x: pd.Series(parse_linestring(x)))

Step 4: Verify the Result #

Running the code above will give you exactly the format you requested:

print(df)
# Output:
#        col1                                                col2
# 0  LINESTRING  [(174.76, -36.8), (174.76, -36.8), (175.76, -36.82)]

Customization Tips #

• To change how values are processed, just modify the process_coord_value function (e.g., add 1 to each value, convert to integers, etc.).
• If your data has other geometry types (like POLYGON), you can adjust the regex or add conditional logic to handle them.

内容的提问来源于stack exchange，提问作者NZ_DJ