Must require Explorer.Dataframe in each cell

[lehoff]

Environment

• Elixir & Erlang/OTP versions (elixir --version): 1.15.2
• Operating system: MacOS Ventura 13.5
• How have you started Livebook (mix phx.server, livebook CLI, Docker, etc): Desktop app
• Livebook version (use git rev-parse HEAD if running with mix): 0.10.0
• Browsers that reproduce this bug (the more the merrier): Orion
• Include what is logged in the browser console: na
• Include what is logged to the server console: na

Current behavior

Include detailed steps to reproduce error, errors and stacktraces if appropriate.

First cell:

require Explorer.DataFrame, as: DF

Second cell:

df = Explorer.Datasets.fossil_fuels()
grouped = DF.group_by(df, ["country"])
DF.mutate(df, new_column: solid_fuel + cement)
grouped
 |> DF.summarise(max_per_capita: max(per_capita))
 |> DF.arrange(desc: max_per_capita)

error: undefined variable "solid_fuel"
  esl/budget/budget_tool/mvp.livemd#cell:xfpznyqojik7ad6hwmtdpt5kjqmjkyd5:5

Expected behavior

A short description on how you expect the code to behave.

I did a require Explorer.Dataframe very early in my Livebook, so I would have expected that the code did what it does in the documentation in every cell, but it only works if I require Explorer.Dataframe in every cell.

Now I know the limitation so I can work around it, but it is not intuitive.

[josevalim]

I cannot reproduce this. Can you share your notebook? :)

[josevalim]

For example, this notebook worked fine here:

# Untitled notebook

```elixir
Mix.install([
  {:explorer, "~> 0.6.1"}
])
```

## Section

```elixir
require Explorer.DataFrame, as: DF
```

```elixir
df = Explorer.Datasets.fossil_fuels()
grouped = DF.group_by(df, ["country"])
DF.mutate(df, new_column: solid_fuel + cement)

grouped
|> DF.summarise(max_per_capita: max(per_capita))
|> DF.arrange(desc: max_per_capita)
```

[cristineguadelupe]

Same here. I was unable to reproduce it and the code example worked without problems.

[lehoff]

I have found the root cause, but I don't know why it happens. I tried to trim down my notebook, but I forgot to try it out before reporting - I will remember that next time, sorry about wasting your time.

Anyway, here is my notebook:

# Untitled notebook

```elixir
Mix.install([
  {:timex, "~> 3.7"},
  {:explorer, "~> 0.6.0"},
  {:kino, "~> 0.9.0"},
  {:kino_explorer, "~> 0.1.4"}
])

Section

use Timex
require Explorer.DataFrame, as: DF

df = Explorer.Datasets.fossil_fuels()
grouped = DF.group_by(df, ["country"])
DF.mutate(df, new_column: solid_fuel + cement)

grouped
|> DF.summarise(max_per_capita: max(per_capita))
|> DF.arrange(desc: max_per_capita)

This is reproducible at my end.
The problem comes when use Timex is in the the same cell (order does not matter). If I split out the use and the require in two separate cells it works just fine.