compare steps

compare_steps.py

import pandas as pd

df_day1 = pd.read_csv("day1_processed.csv")
df_day2 = pd.read_csv("day2_processed.csv")

# Each has columns like: ["relative_time_s", "CPU_usage", "FPS", ...].

# We assume day1["relative_time_s"]=0 at the start of Day 1's test, day2["relative_time_s"]=0 at start of Day 2's test, etc.
# Hypothetical times in seconds from the start of each day's recording
STEP_TIMES = {
    "Step1": {
        "Day1": (300, 480),   # Step 1 goes from 5 min (300s) to 8 min (480s) on Day 1
        "Day2": (600, 780),   # Step 1 goes from 10 min (600s) to 13 min (780s) on Day 2
    },
    "Step2": {
        "Day1": (480, 600),   # maybe step 2 picks up immediately after step 1 on Day 1
        "Day2": (780, 900),
    }
}

# function for extracting a step from a day
def extract_step_data(df, step_name, day_name, start_end):
    start_s, end_s = start_end
    # 1. Subset rows for the step time range
    subset = df[(df["relative_time_s"] >= start_s) & (df["relative_time_s"] <= end_s)].copy()
    
    # 2. Shift time so step starts at t=0
    subset["step_time_s"] = subset["relative_time_s"] - start_s
    
    # 3. Add metadata
    subset["step"] = step_name
    subset["day"] = day_name
    
    return subset

# use function to combine extracted step from multiple days
all_steps_data = []

for step_name, days_dict in STEP_TIMES.items():
    for day_name, (start_s, end_s) in days_dict.items():
        if day_name == "Day1":
            step_data = extract_step_data(df_day1, step_name, day_name, (start_s, end_s))
        elif day_name == "Day2":
            step_data = extract_step_data(df_day2, step_name, day_name, (start_s, end_s))
        # if more days, handle them similarly...

        all_steps_data.append(step_data)

# Combine them into one DataFrame
df_steps_combined = pd.concat(all_steps_data, ignore_index=True)

# now df_steps_combined has data like:
# ["relative_time_s", "step_time_s", "CPU_usage", "FPS", "step", "day", ...]

# example plot
import seaborn as sns

# Filter to Step1 only:
df_step1 = df_steps_combined[df_steps_combined["step"] == "Step1"]

# Simple lineplot with Seaborn, CPU_usage vs. step_time_s, color by day:
sns.lineplot(data=df_step1, x="step_time_s", y="CPU_usage", hue="day")