OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Data & Analytics

Surpriver

Surpriver is a Python command-line stock-analysis application that uses machine learning and anomaly detection to examine price and volume data, generate technical features, and rank stocks by unusual patterns. Its repository documents Yahoo Finance data loading, reusable data dictionaries, CLI or JSON results, and a historical testing mode.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Surpriver — OpenViST project profile in AI Data & Analytics.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Surpriver when you want a repository-based command-line workflow for scanning a stock list or retrospectively examining anomaly signals, and can work with Python dependencies and market-data loading.

Who it is for

  • Stock researchers who want to screen a chosen list of stocks for unusual price-and-volume behavior.
  • Python users interested in inspecting or adapting a command-line anomaly-detection workflow.
  • Analysts who want to compare historical signals with later market bars.

Documented uses

  • Rank stocks by unusual price-and-volume patterns over a selected history window.
  • Save loaded market data in a dictionary for reuse across analyses.
  • Retrospectively compare anomaly predictions with reserved future bars.
  • Inspect or modify the separate data-loading, feature-generation and detection modules.
Before you choose

Requirements & limits

The repository is public and identifies GPL-3.0 as its license. The README documents Python 3, dependency installation and Docker use.

  • Market data is loaded through the documented Yahoo Finance module; the README also supports reusing saved data dictionaries.
  • The tool ranks anomalies rather than confirming future price moves; the README says lower, negative scores are more unusual.
  • The documented output can be printed in the CLI or saved as JSON.
  • Historical testing uses earlier data for predictions and reserves future bars for comparison; the README says the reserved count should exceed five.
  • The repository documents both a Python/dependency setup and a Docker workflow.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run a documented anomaly scan from the command line

Rank stocks by unusual patterns using the README’s example settings.

Before you start

  • Python 3 environment with the repository and its listed dependencies installed from requirements.txt.
  • A configured stock list; the README identifies a stocks location for the symbols to analyze.
  • Access to the Yahoo Finance data-loading workflow, or a previously saved data dictionary.

Steps

  1. Install the listed packages with the command below.

    Documented command
    pip install -r requirements.txt

  2. Run detection_engine.py with the README’s example settings: top_n 25, min_volume 5000, data_granularity_minutes 60, history_to_use 14, is_load_from_dictionary 0, data_dictionary_path dictionaries/data_dict.npy, is_save_dictionary 1, is_test 0 and future_bars 0.

  3. Review the ranked results in the CLI; the documented output format can also be set to JSON.

Expected outcome

A ranked list of the most anomalous stocks for the selected analysis window, with data saved to the specified dictionary path when saving is enabled.

Guide limitations

  • The example uses 60-minute bars and ranks 25 stocks; settings and results depend on the selected symbols and market data.
  • Anomaly scores identify unusual patterns and do not establish that a stock will move substantially.
  • The README lists available granularities as 1, 5, 15, 30 and 60 minutes.
Documentation 1 ↗
Source-based practical guide

Evaluate anomaly signals against historical bars

Use the documented testing mode to compare predictions with reserved future data.

Before you start

  • Python 3 environment with the repository and its listed dependencies installed.
  • A configured stock list and access to the Yahoo Finance data-loading workflow, or a previously saved data dictionary.

Steps

  1. Set is_test to 1 and future_bars to a value greater than 5, as specified in the README.

  2. Run the detection engine with the README’s example settings, including top_n 25, min_volume 5000, 60-minute granularity and history_to_use 14.

  3. Review the resulting historical comparison and figures described by the README.

Expected outcome

A retrospective assessment that compares anomalous-stock predictions with the reserved later bars.

Guide limitations

  • This is historical evaluation, not a guarantee of future performance.
  • The supplied README text documents the example command and test parameters but does not provide a complete description of every output or evaluation metric.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
tradytics/surpriver · Official homepage title · retained in its source language
How we review sourcesSuggest a correction