{"id":20522277,"url":"https://github.com/stevenrice99/llm-forecast","last_synced_at":"2026-04-22T10:02:21.557Z","repository":{"id":250951016,"uuid":"817877414","full_name":"StevenRice99/LLM-Forecast","owner":"StevenRice99","description":"A Novel Hybridized Forecasting Technique Utilizing ARIMA and Large Language Models","archived":false,"fork":false,"pushed_at":"2024-12-20T20:32:36.000Z","size":129486,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-03-06T00:34:07.121Z","etag":null,"topics":["arima","arima-forecasting","arima-model","arima-modeling","forecast","forecasting","forecasting-model","forecasting-models","large-language-model","large-language-models","llama","llama3-1","llm","prediction-model","predictive-analytics","predictive-modeling","prompt-engineering","python","python3","statsmodels"],"latest_commit_sha":null,"homepage":"https://stevenrice.ca#llm-forecast","language":"Python","has_issues":false,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/StevenRice99.png","metadata":{"files":{"readme":"readme.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-06-20T16:13:53.000Z","updated_at":"2025-02-21T16:55:13.000Z","dependencies_parsed_at":"2024-11-15T22:34:57.184Z","dependency_job_id":"9474b870-2188-4c60-905a-0b73e0a125b5","html_url":"https://github.com/StevenRice99/LLM-Forecast","commit_stats":null,"previous_names":["stevenrice99/llm-forecast"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/StevenRice99/LLM-Forecast","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/StevenRice99%2FLLM-Forecast","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/StevenRice99%2FLLM-Forecast/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/StevenRice99%2FLLM-Forecast/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/StevenRice99%2FLLM-Forecast/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/StevenRice99","download_url":"https://codeload.github.com/StevenRice99/LLM-Forecast/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/StevenRice99%2FLLM-Forecast/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32130776,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-22T08:34:57.708Z","status":"ssl_error","status_checked_at":"2026-04-22T08:34:55.583Z","response_time":58,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["arima","arima-forecasting","arima-model","arima-modeling","forecast","forecasting","forecasting-model","forecasting-models","large-language-model","large-language-models","llama","llama3-1","llm","prediction-model","predictive-analytics","predictive-modeling","prompt-engineering","python","python3","statsmodels"],"created_at":"2024-11-15T22:34:47.420Z","updated_at":"2026-04-22T10:02:21.500Z","avatar_url":"https://github.com/StevenRice99.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# A Novel Hybridized Forecasting Technique Utilizing ARIMA and Large Language Models\n\n## Usage\n\n1. This project requires [Ollama](https://ollama.com \"Ollama\"), so ensure it is installed and running.\n2. For web scraping, you need to have [Mozilla Firefox](https://www.mozilla.org/en-CA/firefox \"Mozilla Firefox\") installed.\n3. Install requirements by running ``pip install -r requirements.txt`` in the directory of this project. It is recommended you do this in a [virtual environment](https://docs.python.org/3/tutorial/venv.html \"Python Virtual Environments and Packages\").\n4. To replicate our tests, first delete the ``Data``, ``Results``, and ``Responses`` folders.\n5. Run ``main.py`` to perform the experiments. Sometimes, during the web scraping [Selenium](https://www.selenium.dev \"Selenium\") using [Mozilla Firefox](https://www.mozilla.org/en-CA/firefox \"Mozilla Firefox\") may hang, and not move onto the next news article. In this case, simply restart the script, and it will continue from where it left off.\n   1. ``-f`` or ``--forecast`` - Number of weeks to forecast. Defaults to twelve.\n   2. ``-w`` or ``--width`` - The width of the figures. Defaults to eight.\n   3. ``-t`` or ``--height`` - The height of the figures. Defaults to 3.45.\n   4. ``-d`` or ``--decimals`` - The number of decimal spaces. Defaults to two.\n   5. ``-a`` or ``--alpha`` - The alpha factor for the desired confidence level which by default is 95%.\n   6. ``-c`` or ``--clamp`` - By how much should forecast values be clamped around the baseline prediction. Defaults to one hundred.\n   7. ``-l`` or ``--latest`` - Up to how many latest weeks of data should we keep. Defaults to zero meaning keep all data.\n6. Under the ``Data`` folder, you will see the summaries of the news articles produced by the large language model. Under the ``Results`` folder, you will see the results charts and plots.\n   1. ``Actual.csv`` - The actual COVID-19 hospitalizations to occur over the next given weeks from a given week.\n   2. ``Baseline.csv``, ``Unmasked.csv``, and ``Masked.csv`` - The baseline model and full model predictions of how many COVID-19 hospitalizations to occur over the next given weeks from a given week.\n   3. ``Difference Baseline.csv`` and ``Difference Full Model.csv`` - The difference between the actual results and each of the model results.\n   4. ``Success Rate.csv`` - The success rate of each model, where success was determined if a forecast met or exceeded the actual amounts of hospitalizations to occur over a period.\n   5. ``Average Difference.csv`` - The average difference each model had from the actual amounts of hospitalizations to occur over a given period.\n   6. ``Total Failures.csv`` - The total failures which occurred for each model over a forecasting period.\n   7. ``Total Excess.csv`` - The total excess which occurred for each model over a forecasting period.\n   8. ``WIS.csv`` - The weighted interval scores.\n   9. ``MAE.csv`` - The mean absolute errors.\n   10. ``Coverage.csv`` - Coverage of values falling within the alpha prediction intervals.\n7. ``Terms.txt`` included all terms for COVID-19 which should be masked.\n8. ``Trusted.txt`` includes all trusted publishers.\n   1. To list all untrusted publishers in all the summarized news articles, run ``publishers.py``. Passing in either ``-t`` or ``--trusted`` will list all trusted publishers in all the summarized news articles.\n\n## Data\n\nThe data used for this experiment is taken from [Covid Timeline Canada](https://github.com/ccodwg/CovidTimelineCanada/blob/main/data/pt/hosp_admissions_pt.csv \"Covid Timeline Canada GitHub\").","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fstevenrice99%2Fllm-forecast","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fstevenrice99%2Fllm-forecast","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fstevenrice99%2Fllm-forecast/lists"}