Use # pip install -r requirements.txt on ide terminal you use.
You need python or anaconda environment.

You can see notebook as Last_version.

For use project just open Project as folder then click app.py then use pip install -r requirements.txt on IDE's terminal incase you dont have these libraries.

After Run it, there is local host like http://127.0.0.1:5000/  click on it then there will be opened web page, then enter values then hit the predict button on the right result will show up.

There are predit_with_normalize_less.csv , predictwithnormalize.csv and lessinputs.csv for result of predict on validation dataset. Compare DNI with Label column. DNI target and Label is predict.

We used 2017.csv as dataset.

I used pycaret, automatized ML library for this project.
-------------------------------------------------------------------------------------------------------------------------------------------------------------------

## ABOUT
In these days of global energy crisis, solar energy is emerging as a viable alternative to fossil fuels. 
Even if they are closer to the Equator or receive more sunlight, developing or underdeveloped countries cannot benefit as much from solar energy as developed countries. Solar energy panels are installed and adjusted according to the angle of direct normal irradiance to achieve efficiency. 
However, direct normal irradiance is measured using a pyrheliometer. The pyrheliometer is a sensor that requires ongoing maintenance (calibration) and is more expensive than other sensors. This study attempted to predict direct normal radiation with machine learning models on the dataset created with meteorological data obtained from other sensors without using a pyrheliometer.

[FULL TEXT](https://github.com/elandil2/Predict-Direct-Normal-Irradiance/blob/main/A%20SYSTEM%20TO%20PREDICT%20SOLAR%20RADIATION%20BY%20USING_0722.pdf)