This repository contains the auto-generated sources that we use in nimiq.com.

We are using Github Actions It will run every Monday, Thursday and Saturday at 03:00 UTC.

Requirements

You should have the following environmental variables:

• GITHUB_TOKEN: Read docs.
• TWITTER_TOKEN: Read docs.
• LUNARCRUSH_TOKEN: Read docs.

Run the scripts

python3 ./src/tweets.py
python3 ./src/stats.py
python3 ./src/social_score.py

What is it being generated?

Tweets

We like to know what the community is saying about Nimiq.

• tweets.py
• Tweets job

GitHub Action

• Tweets job

How do we do it?

We use Twitter's API to get the latest tweets that contain the word nimiq.

Then, we filter the tweets using finiteautomata/bertweet-base-sentiment-analysis model from Hugging Face 🤗.

Output

We store two tweets datasets: All tweets and Positive tweets.

GitHub Stats

Compute the amount of commits and additions made in the last N_WEEKS.

• stats.py
• Stats job

How do we do it?

We use GitHub's statistics API to get the stats of the last year, and then we filter the data to get the stats of the last N_WEEKS.

Output

We store two files: Stats and Stats by repo.

Social score by LunarCrush

Fetchs social stats from LunarCrush.

• social_score.py
• Social score job

How do we do it?

We use Lunarcrush's API to get the stats of NIM. These are the stats we are using:

• social_score: Sum of followers, retweets, likes, reddit karma... of social posts collected
• social_score_24h_rank: Position/rank of the output 24 hour social score relative to all other supported output, lower is best/highest social score
• average_sentiment: Average sentiment of collected social posts
• sentiment_absolute: Percent of bullish or very bullish tweets
• sentiment_relative: Percent tweets that are bullish (excluding neutral in the count)
• social_impact_score: A proprietary score based on the relative trend of social_score
• galaxy_score: A proprietary score based on technical indicators of price, average social sentiment, relative social activity
• social_contributors: The number of unique accounts posting on social
• social_volume_calc_24h: Number of social posts over the last 24 hours
• social_score_calc_24h: Sum of social engagement over the last 24 hours

Output

We store the stats in social-score.json.