https://github.com/anselmoo/csv_first_insight
A sklearn-based correlation- and prediction-maker for small *csv-data
https://github.com/anselmoo/csv_first_insight
cmdline csv machine-learning sklearn
Last synced: 7 months ago
JSON representation
A sklearn-based correlation- and prediction-maker for small *csv-data
- Host: GitHub
- URL: https://github.com/anselmoo/csv_first_insight
- Owner: Anselmoo
- License: mit
- Created: 2019-09-10T16:01:20.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2020-08-14T17:27:29.000Z (over 5 years ago)
- Last Synced: 2025-09-01T06:31:26.970Z (7 months ago)
- Topics: cmdline, csv, machine-learning, sklearn
- Language: Python
- Size: 171 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
[](https://travis-ci.com/Anselmoo/csv_first_insight)
[](https://www.codefactor.io/repository/github/anselmoo/csv_first_insight)
[](https://codebeat.co/projects/github-com-anselmoo-csv_first_insight-master)
[](https://github.com/Anselmoo/csv_first_insight/commits/master)
[](https://zenodo.org/badge/latestdoi/207603311)
[](https://github.com/Anselmoo/csv_first_insight/blob/master/LICENSE)
[](https://github.com/Anselmoo/csv_first_insight/releases)
# CSV-First-Insights
A [`sklearn`](https://scikit-learn.org/stable/index.html)-based *correlation- and prediction-maker* for small csv-data < 10,000 entries. Consquently, *no* Neural Network will be used and so far the following Models are implemented:
- [Ridge-Regression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Ridge.html)
- [Gradient-Boosting-Trees](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)
- [Random-Forest-Regressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html)
Furthermore, for a first analysis, the cluster- and *aprori*-pair-plots can be easily generated for checking dependencies in the data.
The **CSV-First-Insights**-application can be installed like this:
python setup.py install
The options of the Command Line Interface of **CSV-First-Insights** are:
python -m pyinsights --help
usage: __main__.py [-h] [--fname FNAME FNAME] [--mode MODE] [--export]
Analyzer for small (# < 10,000) csv-Databases with binary content via scikit-learn!
Training-Set and Test-Set is separately stored in two databases.
optional arguments:
-h, --help show this help message and exit
--fname FNAME FNAME Two filenames have to be defined for the train- and test-set.
Default names are: train-data.csv','test-data.csv'
--mode MODE Please chose the model for the forecaset:
*Ridge-Regression as a Variation of Linear-Regressions -> rig(deafault)
*Gradient-Boosting-Trees -> grad
*Random-Forest -> fors
*All three models, please choose -> all
--export Export the Apriori-Analysis, Cluster-Maps, and Predictions as png- and txt-file
The **CSV-First-Insights** can be also loaded as packages like this:
```python
import pyinsights
import pyinsights.dataread as dr
import pyinsights.mlmodels as ml
import pyinsights.sklsetups as skl
```
The Ridge-Regression-Prediction of **CSV-First-Insights** for the [The Ultimate Halloween Candy Power Ranking](https://www.kaggle.com/fivethirtyeight/the-ultimate-halloween-candy-power-ranking) of kaggle:
