Join the waitlist

← Back to the steam-room data question · Front page

Your data, first look

Client name: Bill

What does my steam room data look like?

Presenting Question

  • Question Verbatim from You: “What does my steam room data look like?”
  • Interpretation of the Question (an attempt to make the question slightly more precise, computable and scientific): What are the characteristics of the steam-room record taken from your conversations with the PerSci chatbotThe PerSci chatbot is a text-message assistant you log your day to in plain words; a session mentioned there is labelled by hand before any analysis uses it. More at persci.app? What are the time range(s)? The frequency of data obtained over time? The quantities reported? Any problematic issues? The objective is not to answer any particular focused question but to characterize what is in the data, and to let you verify — sanity check — that the data imported matches what you expected to see.

Answer to the Presenting Question: 29 sessions, 2026

  • Time range: 29 sessions from 2026-01-07 to 2026-10-09.
  • Gaps: the longest stretch with no session was 47 days. A week without a session means none was mentioned, not that none happened.
  • Frequency: 3.2 sessions a month on average.
  • Quantities: for each session, the day; the time (9 of them a clock time you gave); the minutes (27 of 29, 19 of them exactly 10); and the gym (15 of 29).
  • Problems found: 10 — 1 high, 3 medium, 6 low. Listed at the foot of the page.

Do you have a similar question? Join the waitlist.

Questions asked of this data

This page only describes the record. These each answer one question from it.

When is there data?

Steam-room sessions per week, from the first labelled session to the last. A week with no bar is a week no message mentioned one — not, by itself, a week without one. Hover a bar for its week.

00.511.52sessionsApr 26JulOct
MonthSessions
January 20264
February 20262
March 20261
April 20261
May 20264
June 20264
July 20261
August 20266
September 20265
October 20261

Data Source(s) and Date(s)

  • Data sources:
    • Devices: none. Nothing measured these sessions.
    • Textual sources: your messages to the PerSci chatbotThe PerSci chatbot is a text-message assistant you log your day to in plain words; a session mentioned there is labelled by hand before any analysis uses it. More at persci.app that mention the steam room or a sauna, each labelled by hand, the last on 2026-10-25: whether you steamed, for how long, when and where.
  • Dates: 2026-01-07 to 2026-10-09 — 276 days, 29 sessions
  • People and Organizations Named or Anonymized:
    • Named (in your template names, your notes, or what you told us): Equinox
    • Anonymized (shown by alias, inside your own notes and quotes too): None

Time range(s) and gaps

The record runs from 2026-01-07 to 2026-10-09: the first and the last labelled session. Your chat record itself starts 2025-12-18 and, when these labels were made, ran to 2026-10-23; days outside the series are not in it at all.

Half the time the next session came within 6 days. The five longest stretches between two sessions:

  • 2026-03-13 to 2026-04-29: 47 days
  • 2026-06-29 to 2026-07-30: 31 days
  • 2026-04-29 to 2026-05-20: 21 days
  • 2026-02-21 to 2026-03-13: 20 days
  • 2026-01-13 to 2026-01-27: 14 days

Frequency: how often, and at what time of day

Which day of the week

DaySessions
Monday4
Tuesday5
Wednesday6
Thursday4
Friday2
Saturday5
Sunday3

What time of day

weekdayweekend

8 AMnoon4 PM8 PMtime of dayApr 26JulOct

Most times are when you sent the message, not a clock reading; see “Quantities reported”.

Quantities reported

One row per calendar day, 276 rows, 12 columns. The columns below are counted over the steam days only. Nothing else is recorded: not the temperature, not which steam room, not who was there, and no text from your messages.

ColumnWhat it isSteam days with a value
local_datethe calendar day, Pacific time29 of 29
steamed1 when a message said you steamed; 0 when one said you did not; blank when no message said either29 of 29
duration_minminutes in the steam room, as you said them (two sessions on one day are added together)27 of 29
duration_basiswhether the minutes were said outright or hedged (“about 10”)27 of 29
first_steam_hhmmthe time of the session, to the minute29 of 29
first_steam_minutesthe same time as minutes after midnight29 of 29
time_basishow that time was arrived at — see “How each time was arrived at”29 of 29
locationthe gym, as you typed it15 of 29
n_sessionssessions that day29 of 29
n_sessions_with_durationof those, how many say how long29 of 29
assumednever set in this file: a blank day is left blank, not filled in29 of 29
message_row_idswhich chat messages the row came from (internal numbers, no text)29 of 29

How long

27 of 29 sessions say how long.

MinutesSessions
6 minutes1
8 minutes1
9 minutes1
10 minutes19
11 minutes3
12 minutes1
15 minutes1

How each time was arrived at

The time is…Sessions
when you sent the message (“just did 10 minutes”)18
a clock time you gave9
worked back from “an hour ago”2

Where

15 of 29 sessions name a gym.

Gym, as typedSessions
Equinox Van Mission10
Equinox3
Equinox Union Street1
Equinox Pine St.1

Browse the data

Every session in the record, oldest first. This is the labelled record, not a verdict: check it against what you remember.

DateDayTimeThe time is…MinutesGym
2026-01-07Wed5:34 PMwhen you sent the message (“just did 10 minutes”)10Equinox Van Mission
2026-01-13Tue7:03 PMwhen you sent the message (“just did 10 minutes”)15not named
2026-01-27Tue6:40 PMa clock time you gave10not named
2026-01-31Sat1:15 PMwhen you sent the message (“just did 10 minutes”)10not named
2026-02-10Tue6:51 PMwhen you sent the message (“just did 10 minutes”)10not named
2026-02-21Sat11:53 AMwhen you sent the message (“just did 10 minutes”)10not named
2026-03-13Fri3:48 PMworked back from “an hour ago”10not named
2026-04-29Wed9:44 AMwhen you sent the message (“just did 10 minutes”)10Equinox Van Mission
2026-05-20Wed8:56 AMwhen you sent the message (“just did 10 minutes”)8Equinox Van Mission
2026-05-23Sat10:55 AMwhen you sent the message (“just did 10 minutes”)6Equinox
2026-05-25Mon9:21 AMwhen you sent the message (“just did 10 minutes”)10Equinox
2026-05-27Wed9:09 AMwhen you sent the message (“just did 10 minutes”)10not named
2026-06-07Sun3:58 PMwhen you sent the message (“just did 10 minutes”)11Equinox Union Street
2026-06-11Thu10:51 AMwhen you sent the message (“just did 10 minutes”)10Equinox Van Mission
2026-06-18Thu3:22 PMwhen you sent the message (“just did 10 minutes”)10not named
2026-06-29Mon10:49 AMwhen you sent the message (“just did 10 minutes”)12Equinox Van Mission
2026-07-30Thu4:04 PMworked back from “an hour ago”10Equinox
2026-08-01Sat4:30 PMa clock time you gave10not named
2026-08-03Mon11:51 AMwhen you sent the message (“just did 10 minutes”)10Equinox Pine St.
2026-08-13Thu10:34 AMwhen you sent the message (“just did 10 minutes”)10Equinox Van Mission
2026-08-18Tue1:15 PMa clock time you gave—Equinox Van Mission
2026-08-26Wed4:39 PMwhen you sent the message (“just did 10 minutes”)10Equinox Van Mission
2026-08-30Sun5:01 PMa clock time you gave11not named
2026-09-02Wed1:42 PMwhen you sent the message (“just did 10 minutes”)11Equinox Van Mission
2026-09-06Sun6:20 PMa clock time you gave10not named
2026-09-08Tue7:27 PMa clock time you gave10not named
2026-09-14Mon7:50 PMa clock time you gave10Equinox Van Mission
2026-09-26Sat3:48 PMa clock time you gave—not named
2026-10-09Fri9:51 AMa clock time you gave9not named

Samples of Raw User Utterance Data

These 10 of the 68 messages the labels were made from are meant to be a representative sample. Not fully random, and not picked by hand for being clear or interesting: drawn at random, subject to these constraints:

  • never a message screened out for personal material (see below);
  • at least one of each kind of message;
  • the rest in proportion to how common each kind is;
  • within a kind, spread across the whole record: its messages, in date order, are cut into as many equal stretches as the kind gets samples, and one is drawn at random from each stretch.

Each is quoted exactly as typed, voice-typing errors included.

  • 2026-01-13 Tue, 7:02 PM · a steam-room diary entry that says how long · labelled as: a session, 15 minutes
    “Steam room diary spent a solid 15 minutes in the steam room today just a few minutes ago”
  • 2026-02-01 Sun, 5:32 PM · an activity log listing the steam room among other things · labelled as: not counted as a session
    “Log activity steam room, followed by get dressed, followed by walk to car”
  • 2026-02-28 Sat, 12:21 PM · an activity log listing the steam room among other things · labelled as: not counted as a session
    “Log activity steam room, and shower rinse off at Equinox”
  • 2026-03-24 Tue, 5:48 PM · a diary entry saying you did not steam · labelled as: a day without steam
    “Diary steam room I did not use the steam room today :-)”
  • 2026-05-21 Thu, 10:34 PM · another diary entry (sleep, mood, knees, workout) mentioning it · labelled as: not counted as a session
    “Diary sleep had one glass of red wine at about 7:45 PM no work out no steam room today”
  • 2026-05-27 Wed, 9:09 AM · a steam-room diary entry that says how long · labelled as: a session, 10 minutes
    “Diary steam room sat in the steam room for 10 minutes”
  • 2026-05-27 Wed, 4:27 PM · a command or a note to the chatbot · labelled as: not counted as a session
    “Show diary, steam room”
  • 2026-06-29 Mon, 10:48 AM · an activity log listing the steam room among other things · labelled as: not counted as a session
    “Log activity steam room, and shower”
  • 2026-08-18 Tue, 1:49 PM · a steam-room diary entry that does not say how long · labelled as: a session, length not said
    “Diary steam room finished steam room at Equinox Van Mission at 1:15 today”
  • 2026-09-08 Tue, 7:32 PM · a steam-room diary entry that says how long · labelled as: a session, 10 minutes
    “Diary steamer just finished the steam room about five minutes ago for 10 minutes 10 minutes duration and this time unlike almost all times I actually wore my aura ring throughout”

Which kinds, and what the samples leave out

14 of the 68 messages are never used as samples, because they contain personal material: health or mood, family matters, private situations, or another person named or described. They are still in the counts and the labels. For 2 kinds it is more than a third of the kind, so the samples of those kinds look tidier than the messages themselves, which are often longer and about more than one thing:

  • an activity log listing the steam room among other things
  • another diary entry (sleep, mood, knees, workout) mentioning it

Kind of messageMessagesShownNever shown
a steam-room diary entry that says how long2731
a steam-room diary entry that does not say how long210
a diary entry saying you did not steam110
an activity log listing the steam room among other things2238 — more than a third of this kind
a command or a note to the chatbot510
another diary entry (sleep, mood, knees, workout) mentioning it1115 — more than a third of this kind

Problematic issues

  • high — 5 days whose messages show a session, or might, have none in the record: 2026-02-01, 2026-02-07, 2026-02-28, 2026-03-11, 2026-05-08. The first labelling round marked every “Log activity” entry as not about a session.
    What to do: Relabel those messages. Until then the sleep analysis leaves these days out rather than count them as days without steam.
  • medium — 1 more day with an activity log mentioning the steam room but no diary entry, found by the follow-up question: 2026-07-25. Not among the days the sleep analysis leaves out.
    What to do: Relabel with the five above; until then it counts as a day without steam.
  • medium — A session logged without the words “steam” or “sauna” (“Log activity gym, shower, drive home”) never reached the labelling, so it cannot show up even as missing.
    What to do: Nothing to do from the data; say so if you often log it that way.
  • medium — Only 1 day says outright that you did not steam (2026-03-24). Every other day without a session is silent, so “no steam” there is an assumption.
    What to do: A quick “no steam today” on gym days without one would turn guesses into data.
  • low — A session on 2025-12-24 is mentioned before the record starts, and your chat record begins 2025-12-18: nothing earlier exists.
    What to do: None.
  • low — A gym not named is written two ways: “not stated” on 9 sessions (first labelling round) and left blank on 5 (second round).
    What to do: Read both as not named; this page does.
  • low — 2 sessions say nothing about how long: 2026-08-18, 2026-09-26.
    What to do: None; they still count as sessions.
  • low — 19 of 27 durations are exactly 10 minutes, so minutes here are your rounding, not a timer.
    What to do: Use whether you steamed, not for how long.
  • low — 20 of 29 times are not a clock reading but when you sent the message, or worked back from “an hour ago”. Whether a stated time means the start or the end of the session was never fixed.
    What to do: Fine for the day; too rough for “how long before bed”.
  • low — The record ends at the last session, 2026-10-09, although your chat ran to 2026-10-23. Those later days are not in it, even as days without steam.
    What to do: None.

Follow-up questions: Mostly yes

Do activity logs like “Log activity steam room, and shower” have a steam-room diary entry the same day? If not, are we missing steam-room sessions?

  • Question Verbatim from You: “for entries like 6-19 'log activity steam room, and shower' do there tend to be stream room diary entries associated with entries like those? If not, we're missing some steam room usage data, right?”
  • Interpretation: in your question you said “6-19”. We are interpreting that as the activity log of 2026-06-29, the one with those words. “Associated” we take as a steam-room diary entry on the same calendar day.

Mostly yes. Of the 22 activity logs that mention the steam room, on 21 days, 15 days also have a steam-room diary entry — 2026-06-29 among them. There the diary entry is the record, and the log is its twin.

On the other 6 days there is none, so yes: these are steam-room sessions the record is missing, or may be. Some of these logs read as a plan (“… and then do the steam room”) rather than a session done.

  • 5 already listed under Problematic issues: 2026-02-01, 2026-02-07, 2026-02-28, 2026-03-11, 2026-05-08.
  • 1 newly found by this check: 2026-07-25. It is not among the days the sleep analysis leaves out, so there it counts as a day without steam.

One of the extra checks, not a registered test: it counts days, and reads none of the messages' words beyond the opening “Log activity”.

If you have a health-related question about your own data, join the waitlist for your own custom analysis.

Act soon: it is free while we stand this service up and refine it. All we ask in return is permission to publish a results page like the ones we have made for our other clients. We never publish your surname or your date of birth. The page shows a first name or a pseudonym, whichever you prefer, and your age in years and your sex.

Made by PerSci from your steam-room record. It describes your log; it is not medical advice.