Note:

Because this dataset has been used in a competition, we had to hide some of the data to prepare the test dataset for the competition. Thus, in the previous version of the dataset, only train.csv file is existed.

Content

This dataset represents 10 different physical poses that can be used to distinguish 5 exercises. The exercises are Push-up, Pull-up, Sit-up, Jumping Jack and Squat. For every exercise, 2 different classes have been used to represent the terminal positions of that exercise (e.g., “up” and “down” positions for push-ups).

Collection Process

About 500 videos of people doing the exercises have been used in order to collect this data. The videos are from Countix Dataset that contain the YouTube links of several human activity videos. Using a simple Python script, the videos of 5 different physical exercises are downloaded. From every video, at least 2 frames are manually extracted. The extracted frames represent the terminal positions of the exercise.

Processing Data

For every frame, MediaPipe framework is used for applying pose estimation, which detects the human skeleton of the person in the frame. The landmark model in MediaPipe Pose predicts the location of 33 pose landmarks (see figure below). Visit Mediapipe Pose Classification page for more details.

https://mediapipe.dev/images/mobile/pose_tracking_full_body_landmarks.png" alt="33 pose landmarks">

Google data search exercises can be used to practice finding data or statistics on a topic of interest, including using Google's own internal tools and by using advanced operators.

The MEx Multi-modal Exercise dataset contains data of 7 different physiotherapy exercises, performed by 30 subjects recorded with 2 accelerometers, a pressure mat and a depth camera.

Application The dataset can be used for exercise recognition, exercise quality assessment and exercise counting, by developing algorithms for pre-processing, feature extraction, multi-modal sensor fusion, segmentation and classification.

** Data collection method ** Each subject was given a sheet of 7 exercises with instructions to perform the exercise at the beginning of the session. At the beginning of each exercise the researcher demonstrated the exercise to the subject, then the subject performed the exercise for maximum 60 seconds while being recorded with four sensors. During the recording, the researcher did not give any advice or kept count or time to enforce a rhythm.

** Sensors** Obbrec Astra Depth Camera - sampling frequency – 15Hz - frame size – 240x320

Sensing Tex Pressure Mat - sampling frequency – 15Hz - frame size – 32*16

Axivity AX3 3-Axis Logging Accelerometer - sampling frequency – 100Hz - range – 8g

** Sensor Placement** All the exercises were performed lying down on the mat while the subject wearing two accelerometers on the wrist and the thigh. The depth camera was placed above the subject facing down-words recording an aerial view. Top of the depth camera frame was aligned with the top of the pressure mat frame and the subject’s shoulders such that the face will not be included in the depth camera video.

** Data folder ** MEx folder has four folders, one for each sensor. Inside each sensor folder, 30 folders can be found, one for each subject. In each subject folder, 8 files can be found for each exercise with 2 files for exercise 4 as it is performed on two sides. (The user 22 will only have 7 files as they performed the exercise 4 on only one side.) One line in the data files correspond to one timestamped and sensory data.

Attribute Information The 4 columns in the act and acw files is organized as follows: 1 – timestamp 2 – x value 3 – y value 4 – z value Min value = -8 Max value = +8

The 513 columns in the pm file is organized as follows: 1 - timestamp 2-513 – pressure mat data frame (32x16) Min value – 0 Max value – 1

The 193 columns in the dc file is organized as follows: 1 - timestamp 2-193 – depth camera data frame (12x16) dc data frame is scaled down from 240x320 to 12x16 using the OpenCV resize algorithm Min value – 0 Max value – 1

Context

Below are my recorded workouts going back almost 3 years, using the Strong app. Nearly every single movement I have performed in the gym is recorded here with the exception of some warmup sets.

It would be interesting if anyone could find any useful patterns, surprisingly insights, or tips for getting stronger.

Content

Some things to keep in mind when analyzing the data:

• All units are in pounds.
• I generally followed a push-pull-legs-rest-repeat (PPL) split, although not necessarily in that order. As you will see, I followed a generally consistent program over this time but with deviations in reps/sets. Some days I had little energy due to illness, work, lack of sleep, or other events.
• Assume I weight 220LBs for body weight activities - dips, chin-ups, neutral grips chin-ups, and pull-ups. Any value you see for 'weight' in the dataset for these types of exercises can be thought of as added weight. For example, "70" would mean 220+70=290LBs.
• Dumbbell weights are the combined weight of both arms. For example, a dumbbell bench using 100s would be recorded as 200 pounds.
• Some of the same exercises may be entered with different names. For example, "Lateral Raise (Dumbbell)" and "Lateral Raise (Dumbbells)" are the same exercise just entered with different names.
• A proxy for one-rep-max can be found below. This may be useful in comparing strength levels over time from the same exercise but at different rep ranges. I encourage you to look at this from other angles as well. MAX = WEIGHT /(1.0278 - 0.0278*reps)
• Not every set or exercise is performed to the absolute maximum number of reps possible at that time. Sometimes I am fatigued from work, previous sets/workouts, plan on doing a higher volume day, or plan for a slight backoff day. I think this is common sense, but just stating so.
• Any dumbbell movement is generally recorded as the combined weight of both arms. For example, using a dumbbell bent press
• Anything that looks extreme is likely a typo. For example a lift in the 1000s of pounds is a typo.
• Warmups are generally not recorded.
• I wouldn't read too much into workout names as these are generally a close categorization of legs/push/pull.
• I rarely do cardio, but have managed to walk about 2.75 miles per day consistently over this time. Within the last 2 months, I have begun to bike approximately 30 miles per week.

Acknowledgements

Data is exported from the STRONG app, which is a great way to track your workouts.

Inspiration

I hope this dataset is of use to anyone interested in fitness. It would be awesome if this can be of use to others either in improving their own fitness, understanding the complexity of designing routines, and especially if any valuable insights can be generated to improve moving forward.

The raw data on behavior and physical fitness. The behavior for sampling worker before joining WE is on sheet behavior 31 and 62 Then, we show all data for behavior and physical fitness.

This data set is an example data set for the data set used in the experiment of the paper "A Multilevel Analysis and Hybrid Forecasting Algorithm for Long Short-term Step Data". It contains two parts of hourly step data and daily step data

Dataset

This dataset was created by Pawan Saini

Released under CC0: Public Domain

Contents

This table contains 1260 series, with data for years 1990 - 1998 (not all combinations necessarily have data for all years), and was last released on 2007-01-29. This table contains data described by the following dimensions (Not all combinations are available): Geography (30 items: Austria; Belgium (Flemish speaking); Belgium; Belgium (French speaking) ...), Sex (2 items: Males; Females ...), Age group (3 items: 11 years;13 years;15 years ...), Frequency of exercise (7 items: Everyday; Once a week;2 to 3 times a week;4 to 6 times a week ...).

Zip file with 5 Roam exercises in PDF and PPTX formats, plus trainer guide and Quick Guide. GIS vector data. This dataset was first accessioned in the EDINA ShareGeo Open repository on 2014-04-10 and migrated to Edinburgh DataShare on 2017-02-22.

1) Data Introduction • The Gym Members Exercise Dataset is a dataset built to systematically collect gym members' exercise routines, body information, exercise habits, and fitness indicators to analyze individual exercise patterns and health conditions.

2) Data Utilization (1) Gym Members Exercise Dataset has characteristics that: • This dataset contains various body and exercise related numerical and categorical variables such as age, gender, weight, height, body fat percentage, BMI, exercise type (e.g., aerobic, muscular, yoga, HIIT), exercise frequency, session time, heart rate (maximum, average, rest), calorie consumption, water intake, and experience level. (2) Gym Members Exercise Dataset can be used to: • Exercise effect analysis and customized fitness strategy: Various variables such as exercise type, frequency, session time and heart rate, calorie burn, body fat percentage, etc. can be analyzed and used to establish customized exercise plans for each member and optimize exercise effectiveness. • Healthcare and Member Characteristics Based Marketing: Based on demographics and exercise habit data such as age, gender, and experience level, it can be used to develop healthcare programs, segment members, and establish targeted marketing strategies.

This program establishes a deep learning model of CNN+LSTM, which is used for continuous monitoring of exercise heart rate with PPG signals containing motion artifacts, and has achieved good results in the PPG-DaLiA database. The description is as follows: 1. The file main_program_file is the main file, including model construction, data processing, data training, model data verification, and other processing programs for PPG signals that are not used in this article. model: build exercise heart rate monitoring model file; activity_time.xls: Collect each activity time node of each volunteer signal obtained from the PPG-DaLiA database; original_data_read.py: signal data preprocessing program (signal from the PPG-DaLiA database); ppg_filed_hr_cornet_estimate.py: training and prediction program for all volunteers’ PPG signals; ppg_filed_hr_cornet_estimate_single.py: a program to predict the PPG signal of a single volunteer; _1d_cnn, _2d_cnn, ppg_excerise_cnn_type.py, ppg_filed_hr_cnn_estimate.py: programs that use the CNN method for prediction; spc_hr_cornet_estimate.py, spc_hr_cnn_estimate.py: programs for predicting and verifying using other database PPG signals. save_model_estimate_hr.py, save_model_estimate_hr_spc.py: save the heart rate prediction model and the model program for the heart rate prediction model to be used in the SPC database. out_fig: model prediction picture output folder; 2. Data source The data comes from the PPG-DaLiA database (PPG Data For Daily Life Activity, https://archive.ics.uci.edu/ml/datasets/PPG-DaLiA): The database comes from Robert Bosch GmbH and Bosch Sensortec GmbH. The signals in this database come from 15 volunteers of different ages and different physical conditions. PPG and heart rate data are continuously collected during different exercises. The preprocessing of the downloaded data is in the program original_data_read.py. 3.other _0_basic_fun, ch3_preprocess, my_pyhht_lib: some external references of the main program, mainly the functions called by the data preprocessing part, and the main program can view their functions.This program establishes a deep learning model of CNN+LSTM, which is used for continuous monitoring of exercise heart rate with PPG signals containing motion artifacts, and has achieved good results in the PPG-DaLiA database. The description is as follows: 1. The file main_program_file is the main file, including model construction, data processing, data training, model data verification, and other processing programs for PPG signals that are not used in this article. model: build exercise heart rate monitoring model file; activity_time.xls: Collect each activity time node of each volunteer signal obtained from the PPG-DaLiA database; original_data_read.py: signal data preprocessing program (signal from the PPG-DaLiA database); ppg_filed_hr_cornet_estimate.py: training and prediction program for all volunteers’ PPG signals; ppg_filed_hr_cornet_estimate_single.py: a program to predict the PPG signal of a single volunteer; _1d_cnn, _2d_cnn, ppg_excerise_cnn_type.py, ppg_filed_hr_cnn_estimate.py: programs that use the CNN method for prediction; spc_hr_cornet_estimate.py, spc_hr_cnn_estimate.py: programs for predicting and verifying using other database PPG signals. save_model_estimate_hr.py, save_model_estimate_hr_spc.py: save the heart rate prediction model and the model program for the heart rate prediction model to be used in the SPC database. out_fig: model prediction picture output folder; 2. Data source The data comes from the PPG-DaLiA database (PPG Data For Daily Life Activity, https://archive.ics.uci.edu/ml/datasets/PPG-DaLiA): The database comes from Robert Bosch GmbH and Bosch Sensortec GmbH. The signals in this database come from 15 volunteers of different ages and different physical conditions. PPG and heart rate data are continuously collected during different exercises. The preprocessing of the downloaded data is in the program original_data_read.py. 3.other _0_basic_fun, ch3_preprocess, my_pyhht_lib: some external references of the main program, mainly the functions called by the data preprocessing part, and the main program can view their functions.

Includes the raw IMU data for 20 participants performing seven different running exercises each.

20 participants: 16 m, 4 f, 16 to 31 yo, healthy, do sports regularly Seven exercises: Carioca left, carioca right, heel-to-butt, high-knee running, sideskips left, sideskips right, regular running Four IMUs: accelerometer + gyroscope each, two at wrists, two at ankles Ten seconds per recording under supervision One .json file per recording with sensor values and timestamps

A messy data for demonstrating "how to clean data using spreadsheet". This dataset was intentionally formatted to be messy, for the purpose of demonstration. It was collated from here - https://openafrica.net/dataset/historic-and-projected-rainfall-and-runoff-for-4-lake-victoria-sub-regions

In this paper, we introduce a novel benchmarking framework designed specifically for evaluations of data science agents. Our contributions are three-fold. First, we propose DSEval, an evaluation paradigm that enlarges the evaluation scope to the full lifecycle of LLM-based data science agents. We also cover aspects including but not limited to the quality of the derived analytical solutions or machine learning models, as well as potential side effects such as unintentional changes to the original data. Second, we incorporate a novel bootstrapped annotation process letting LLM themselves generate and annotate the benchmarks with ``human in the loop''. A novel language (i.e., DSEAL) has been proposed and the derived four benchmarks have significantly improved the benchmark scalability and coverage, with largely reduced human labor. Third, based on DSEval and the four benchmarks, we conduct a comprehensive evaluation of various data science agents from different aspects. Our findings reveal the common challenges and limitations of the current works, providing useful insights and shedding light on future research on LLM-based data science agents.

This is one of DSEval benchmarks.

UI-PRMD is a data set of movements related to common exercises performed by patients in physical therapy and rehabilitation programs. The data set consists of 10 rehabilitation exercises. A sample of 10 healthy individuals repeated each exercise 10 times in front of two sensory systems for motion capturing: a Vicon optical tracker, and a Kinect camera. The data is presented as positions and angles of the body joints in the skeletal models provided by the Vicon and Kinect mocap systems.

The data presented here was extracted from a larger dataset collected through a collaboration between the Embedded Systems Laboratory (ESL) of the Swiss Federal Institute of Technology in Lausanne (EPFL), Switzerland and the Institute of Sports Sciences of the University of Lausanne (ISSUL). In this dataset, we report the extracted segments used for an analysis of R peak detection algorithms during high intensity exercise.

Protocol of the experiments
The protocol of the experiment was the following.

• 22 subjects performing a cardio-pulmonary maximal exercise test on a cycle ergometer, using a gas mask. A single-lead electrocardiogram (ECG) was measured using the BIOPAC system.
• An initial 3 min of rest were recorded.
• After this baseline, the subjects started cycling at a power of 60W or 90W depending on their fitness level.
• Then, the power of the cycle ergometer was increased by 30W every 3 min till exhaustion (in terms of maximum oxygen uptake or VO2max).
• Finally, physiology experts assessed the so-called ventilatory thresholds and the VO2max based on the pulmonary data (volume of oxygen and CO2).

Description of the extracted dataset

The characteristics of the dataset are the following:

• We report only 20 out of 22 subjects that were used for the analysis, because for two subjects the signals were too corrupted or not complete. Specifically, subjects 5 and 12 were discarded.
• The ECG signal was sampled at 500 Hz and then downsampled at 250 Hz. The original ECG signal were measured at maximum 10 mV. Then, they were scaled down by a factor of 1000, hence the data is represented in uV.
• For each subject, 5 segments of 20 s were extracted from the ECG recordings and chosen based on different phases of the maximal exercise test (i.e., before and after the so-called second ventilatory threshold or VT2, before and in the middle of VO2max, and during the recovery after exhaustion) to represent different intensities of physical activity.

seg1 --> [VT2-50,VT2-30]
seg2 --> [VT2+60,VT2+80]
seg3 --> [VO2max-50,VO2max-30]
seg4 --> [VO2max-10,VO2max+10]
seg5 --> [VO2max+60,VO2max+80]

• The R peak locations were manually annotated in all segments and reviewed by a physician of the Lausanne University Hospital, CHUV. Only segment 5 of subject 9 could not be annotated since there was a problem with the input signal. So, the total number of segments extracted were 20 * 5 - 1 = 99.

Format of the extracted dataset

The dataset is divided in two main folders:

• The folder `ecg_segments/` contains the ECG signals saved in two formats, `.csv` and `.mat`. This folder includes both raw (`ecg_raw`) and processed (`ecg`) signals. The processing consists of a morphological filtering and a relative energy non filtering method to enhance the R peaks. The `.csv` files contain only the signal, while the `.mat` files include the signal, the time vector within the maximal stress test, the sampling frequency and the unit of the signal amplitude (uV, as we mentioned before).
• The folder `manual_annotations/` contains the sample indices of the annotated R peaks in `.csv` format. The annotation was done on the processed signals.

Includes accelerometer data using an ActiGraph to assess usual sedentary, moderate, vigorous, and very vigorous activity at baseline, 6 weeks, and 10 weeks. Includes relative reinforcing value (RRV) data showing how participants rated how much they would want to perform both physical and sedentary activities on a scale of 1-10 at baseline, week 6, and week 10. Includes data on the breakpoint, or Pmax of the RRV, which was the last schedule of reinforcement (i.e. 4, 8, 16, …) completed for the behavior (exercise or sedentary). For both Pmax and RRV score, greater scores indicated a greater reinforcing value, with scores exceeding 1.0 indicating increased exercise reinforcement. Includes questionnaire data regarding preference and tolerance for exercise intensity using the Preference for and Tolerance of Intensity of Exercise Questionnaire (PRETIEQ) and positive and negative outcome expectancy of exercise using the outcome expectancy scale (OES). Includes data on height, weight, and BMI. Includes demographic data such as gender and race/ethnicity. Resources in this dataset:Resource Title: Actigraph activity data. File Name: AGData.csvResource Description: Includes data from Actigraph accelerometer for each participant at baseline, 6 weeks, and 10 weeks.Resource Title: RRV Data. File Name: RRVData.csvResource Description: Includes data from RRV at baseline, 6 weeks, and 10 weeks, OES survey data, PRETIE-Q survey data, and demographic data (gender, weight, height, race, ethnicity, and age).

The PHYTMO database contains data from physical therapy exercises and gait variations recorded with magneto-inertial sensors, including information from an optical reference system. PHYTMO includes the recording of 30 volunteers, aged between 20 and 70 years old. A total amount of 6 exercises and 3 gait variations commonly prescribed in physical therapies were recorded. The volunteers performed two series with a minimum of 8 repetitions in each one. Four magneto-inertial sensors were placed on the lower-or upper-limbs for the recording of the motions together with passive optical reflectors. The files include the specifications of the inertial sensors and the cameras. The database includes magneto-inertial data (linear acceleration, turn rate and magnetic field), together with a highly accurate location and orientation in the 3D space provided by the optical system (errors are lower than 1mm). The database files were stored in CSV format to ensure usability with common data processing software. The main aim of this dataset is the availability of inertial data for two main purposes: the analysis of different techniques for the identification and evaluation of exercises monitored with inertial wearable sensors and the validation of inertial sensor-based algorithms for human motion monitoring that obtains segments orientation in the 3D space. Furthermore, the database stores enough data to train and evaluate Machine Learning-based algorithms. The age range of the participants can be useful for establishing age-based metrics for the exercises evaluation or the study of differences in motions between different aged groups. Finally, the MATLAB function features_extraction, developed by the authors, is also given. This function splits signals using a sliding window, returning its segments, and extract signal features, in the time and frequency domains, based on prior studies of the literature.

This record contains the data files used in exercises in the NBIS course "Introduction to Data Management Practices".

Matlab codes for the solution of exercises which can be found in the book "Tensor Calculus and Differential Geometry for Engineers"

Related publications:

Shahab Sahraee, Peter Wriggers: Tensor Calculus and Differential Geometry for Engineers, Springer, Berlin, to be published.