Bedre med små prototyper enn uferdige store prosjekt.
- Include image from TA powerpoint on training model datapoipeline
- [] Define sudocode for agent interactons
- Make a mini version similar to a vcv rack wire connection.
- build on it
- implement click to add module
- implement doubble click module to define i/o
- [ ]
- build on it
- Make working baseline datafetcher for the websites listed below:
- OECD Data – Data on economy, society, and environment from developed countries
- World Bank Open Data – Extensive economic and development data
- IMF Data – International Monetary Fund economic and financial data
- UNdata – United Nations statistical databases
- Eurostat – European Union statistics
- Trading Economics – Macro indicators and forecasts (limited free API usage)
- Our World in Data – Global development, health, energy, and more
- Quandl / Open Financial Data Project – Various financial and economic datasets (many free)
- and store each in a sqlite .db file.
fetching data into file form single line command python3:
import wbdata, sqlite3, pandas as pd; from datetime import datetime; wbdata.get_series("NY.GDP.PCAP.CD", country="all", date=(datetime(2010,1,1), datetime(2020,1,1)), freq='Y').reset_index().dropna().to_sql("gdp_per_capita", sqlite3.connect("gdp_per_capita.db"), if_exists="replace", index=False)- Good resources for sqlite:
- https://www.youtube.com/watch?v=8Xyn8R9eKB8 Bra video
- /Users/lhh/Documents/LambdaTheSimulationProject/mainProject/ModelProject/DataPipeline/Collection