getdaft (Superseded by Daft)

0.5.0 · renamed · verified Thu Apr 16

The `getdaft` package served as an initial entry point for the Daft distributed dataframe library. As of its `0.5.0` version, it acts primarily as a wrapper that installs an older, specific version of the `daft` package (currently `daft==0.5.0`). Users should directly install and use the actively developed `daft` library instead, as `getdaft` is no longer maintained.

Common errors

```
ModuleNotFoundError: No module named 'getdaft'
```
cause You are trying to import from `getdaft`, but even if `getdaft` is installed, its components are exposed under the `daft` namespace.

fix Change your import statements from `import getdaft` or `from getdaft import ...` to `import daft` or `from daft import ...`.
```
AttributeError: module 'daft' has no attribute 'some_new_feature'
```
cause You have `getdaft` installed, which provides an older version of the `daft` library (e.g., `daft==0.5.0`). The feature you are trying to use was introduced in a newer version of `daft`.

fix Uninstall `getdaft` and any existing `daft` installations, then install the latest `daft` directly: `pip uninstall getdaft daft` followed by `pip install daft`.

Warnings

breaking The `getdaft` package is no longer actively maintained and has been effectively superseded by the `daft` package. New development, features, and bug fixes are exclusively applied to `daft`.
Fix: Migrate your project to use `daft` directly. Uninstall `getdaft` and `pip install daft` to get the latest version.
gotcha Installing `getdaft` (e.g., `0.5.0`) will install `daft==0.5.0`, which is a specific, older version of the `daft` library. This can lead to unexpected behavior or missing features if your code expects a newer `daft` version.
Fix: Always install `daft` directly using `pip install daft` to ensure you get the latest, actively developed version.
gotcha Despite `getdaft` being the installed package name, all core functionalities and classes are exposed under the `daft` namespace (e.g., `from daft import DataFrame`). Attempting to import from `getdaft` will fail.
Fix: Always use `import daft` or `from daft import ...` in your Python code.

Install

pip install getdaft Do NOT use this for new projects, it installs an outdated Daft version.
pip install daft Install the active Daft library (recommended)

Imports

DataFrame
wrong:
```
from getdaft import DataFrame
```
correct:
```
from daft import DataFrame
```
Even when `getdaft` is installed, the actual classes come from the `daft` namespace. Incorrectly trying to import from `getdaft` will result in an `ImportError`.
Daft
wrong:
```
import getdaft
```
correct:
```
import daft
```
The top-level package for operations and builders is `daft`, not `getdaft`.

Quickstart

This quickstart demonstrates usage with the actively developed `daft` library, which is the recommended approach. While `getdaft` installs an older version of `daft`, direct installation of `daft` ensures you are using the latest features and bug fixes. The imports are from the `daft` package namespace.

import daft
import pandas as pd

# IMPORTANT: Use 'pip install daft' directly.
# 'getdaft' installs an old version of 'daft' and should not be used.

# Create a sample Pandas DataFrame
pd_df = pd.DataFrame({
    "name": ["Alice", "Bob", "Charlie"],
    "age": [25, 30, 35]
})

# Convert to Daft DataFrame
daft_df = daft.from_pandas(pd_df)

# Perform a simple operation
result_df = daft_df.select(
dafa_df["name"],
    daft_df["age"] + 1  # Increment age
)

# Collect the result to a Pandas DataFrame for display
print("Resulting Daft DataFrame (collected to Pandas):")
print(result_df.collect().to_pandas())

view raw JSON →