climbing_stats.py

gistfile1.txt

#!/usr/bin/env python3

# climbing_stats.py - Compute climbing stats from Google location history
#
# Author: Amber Sprenkels <amber@electricdusk.com>
# Date: 2024-03-12
#
# This scripts computes the number of times I went to a local climbing or
# bouldering gym close to where I live, based on my Google location history.

import argparse
import json
import datetime
import logging
import os

import polars as pl

STADSMUUR_LOCATION = (51.821852202326895, 5.788025389774361)
GRIP_LOCATION = (51.8130421187338, 5.838377108340926)
EARTH_CIRCUMFERENCE = 40075000  # in meters
CLOSEBY_DISTANCE = 50  # in meters
ACCURATE_DISTANCE = 100  # in meters


if __name__ == "__main__":
    # Set up the logging
    logging.basicConfig(level=os.environ.get("PYTHON_LOG", "INFO").upper())
    logger = logging.getLogger(__name__)

    pl.Config.set_tbl_rows(1000)

    # Parse the arguments and open a file with the climbing stats
    parser = argparse.ArgumentParser(description="Compute climbing stats from Google location history")
    parser.add_argument(
        "takeout", type=argparse.FileType("r"), help="The 'Records.json' location history json file"
    )
    args = parser.parse_args()

    # Read the json file into a DataFrame
    logger.debug(f"reading json file '{args.takeout.name}'")
    json_data = json.load(args.takeout)
    data = {
        "latitude": [],
        "longitude": [],
        "accuracy": [],
        "timestamp": [],
        "source": [],
    }
    for i, loc in enumerate(json_data["locations"]):
        timestamp = loc.get("timestamp")
        if timestamp is not None:
            timestamp = datetime.datetime.fromisoformat(timestamp)
        data["latitude"].append(loc.get("latitudeE7") / 10**7)
        data["longitude"].append(loc.get("longitudeE7") / 10**7)
        data["accuracy"].append(loc.get("accuracy"))
        data["timestamp"].append(timestamp)
        data["source"].append(loc.get("source"))

    logger.debug(f"constructing data frame")
    df = pl.DataFrame(
        data,
        schema={
            "latitude": pl.Float32,
            "longitude": pl.Float32,
            "accuracy": pl.Int32,
            "timestamp": pl.Datetime,
            "source": pl.Utf8,
        },
    )

    # Filter when I was close to stadsmuur or grip using geometrical distance
    # (not taking into account the curvature of the earth)
    is_close_stadsmuur = (pl.col("latitude") - STADSMUUR_LOCATION[0]) ** 2 + (
        pl.col("longitude") - STADSMUUR_LOCATION[1]
    ) ** 2 < (CLOSEBY_DISTANCE / EARTH_CIRCUMFERENCE * 360) ** 2
    is_close_grip = (pl.col("latitude") - GRIP_LOCATION[0]) ** 2 + (
        pl.col("longitude") - GRIP_LOCATION[1]
    ) ** 2 < (CLOSEBY_DISTANCE / EARTH_CIRCUMFERENCE * 360) ** 2
    is_accurate = pl.col("accuracy") < ACCURATE_DISTANCE
    is_close = (is_close_stadsmuur | is_close_grip) & is_accurate
    is_during_day = pl.col("timestamp").dt.hour() >= 8

    # Get instances where I was close to stadsmuur for at least 15 consecutive minutes
    df = (
        df.filter(is_close & is_during_day)
        .with_columns(
            pl.col("timestamp").dt.date().alias("date"),
            is_close_stadsmuur.alias("stadsmuur"),
            is_close_grip.alias("grip"),
        )
        .group_by("date")
        .agg(
            pl.col("timestamp").min().alias("start"),
            pl.col("timestamp").max().alias("end"),
            (pl.col("timestamp").max() - pl.col("timestamp").min()).alias("duration"),
            pl.col("stadsmuur").sum().ne(0).alias("stadsmuur_count"),
            pl.col("grip").sum().ne(0).alias("grip_count"),
        )
        .filter(pl.col("duration") > (15 * 60))
        .sort("date")
    )

    # print(df)
    print(
        df.group_by(pl.col("date").dt.year().alias("year"))
        .sum()
        .sort("year")
        .drop(["date", "start", "end", "duration"])
    )

Today's output:

shape: (9, 3)
┌──────┬─────────────────┬────────────┐
│ year ┆ stadsmuur_count ┆ grip_count │
│ ---  ┆ ---             ┆ ---        │
│ i32  ┆ u32             ┆ u32        │
╞══════╪═════════════════╪════════════╡
│ 2016 ┆ 0               ┆ 3          │
│ 2017 ┆ 0               ┆ 12         │
│ 2018 ┆ 0               ┆ 43         │
│ 2019 ┆ 5               ┆ 19         │
│ 2020 ┆ 7               ┆ 18         │
│ 2021 ┆ 13              ┆ 17         │
│ 2022 ┆ 17              ┆ 15         │
│ 2023 ┆ 22              ┆ 12         │
│ 2024 ┆ 6               ┆ 2          │
└──────┴─────────────────┴────────────┘