Python Client Example

Initialize the client¶

In [ ]:

import crowdcent_challenge as cc
import polars as pl
import numpy as np


client = cc.ChallengeClient(
    challenge_slug="hyperliquid-ranking",
    api_key="hD5EB4pf.KvY8281Kz4igdNBCK7P0RL6tG6J8EaAU",
    base_url="http://localhost:8080/api",
)

import crowdcent_challenge as cc import polars as pl import numpy as np client = cc.ChallengeClient( challenge_slug="hyperliquid-ranking", api_key="hD5EB4pf.KvY8281Kz4igdNBCK7P0RL6tG6J8EaAU", base_url="http://localhost:8080/api", )

Get challenge and training datasets details¶

In [ ]:

client.get_challenge()

client.get_challenge()

In [ ]:

client.list_training_datasets()

client.list_training_datasets()

Download current inference data¶

In [ ]:

client.download_inference_data(
    release_date="current", dest_path="inference_data.parquet"
)

client.download_inference_data( release_date="current", dest_path="inference_data.parquet" )

In [ ]:

# Read in the inference data, in this case, just a universe
inference_data = pl.read_parquet("inference_data.parquet")
inference_data.head()

# Read in the inference data, in this case, just a universe inference_data = pl.read_parquet("inference_data.parquet") inference_data.head()

Submit predictions¶

In [ ]:

# Add prediction columns with random float values for demonstration
# Replace this with your actual prediction logic

inference_data = inference_data.with_columns(
    [
        pl.Series("pred_10d", np.random.random(len(inference_data))).cast(pl.Float64),
        pl.Series("pred_30d", np.random.random(len(inference_data))).cast(pl.Float64),
    ]
)

# You can save this modified dataframe if needed
inference_data.write_parquet("inference_data_with_predictions.parquet")

# Display the first few rows to verify the new columns
print("Inference data with prediction columns:")
inference_data.head()

# Add prediction columns with random float values for demonstration # Replace this with your actual prediction logic inference_data = inference_data.with_columns( [ pl.Series("pred_10d", np.random.random(len(inference_data))).cast(pl.Float64), pl.Series("pred_30d", np.random.random(len(inference_data))).cast(pl.Float64), ] ) # You can save this modified dataframe if needed inference_data.write_parquet("inference_data_with_predictions.parquet") # Display the first few rows to verify the new columns print("Inference data with prediction columns:") inference_data.head()

In [ ]:

client.submit_predictions(file_path="inference_data_with_predictions.parquet", slot=1)

client.submit_predictions(file_path="inference_data_with_predictions.parquet", slot=1)

In [ ]:

for slot in range(1, 6):
    client.submit_predictions(
        file_path="inference_data_with_predictions.parquet", slot=slot
    )

for slot in range(1, 6): client.submit_predictions( file_path="inference_data_with_predictions.parquet", slot=slot )

Meta-model¶

In [ ]:

client.download_meta_model(dest_path="meta_model.parquet")
pl.read_parquet("meta_model.parquet").head()

client.download_meta_model(dest_path="meta_model.parquet") pl.read_parquet("meta_model.parquet").head()