Empty plot when using DataFrame as Chart object

Chart() seems to get a url to dataframe but not a DataFrame object as described in the guide. But this is not handy as, while doing analysis, you usually load data on top, go through some stats and calcualtions and than visulize them. Basically, you cannot use the DataFrame object created with pd.read_csv into Chart(). See example below:

` stocks = pd.read_csv(‘C:\Program Files\Python\Python36-32\lib\site-packages\vega_datasets\data\stocks.csv’)

type(stocks)

stocks.tail()

stocks.groupby([‘symbol’]).agg({‘price’: ‘mean’})

alt.Chart(data=stocks).mark_line().encode( x=‘date:T’, y=‘price:Q’, color=“symbol:N”, column=“symbol:N”) `

[jakevdp]

OK, if that’s the output of to_dict() on your chart, then it looks like you’re not passing the dataframe to as chart data, but rather passing a string containing the local CSV file path to the chart. That local filepath is not a valid URL, which is why you’re not getting any output.

If you want to pass data using a string referencing a local filepath, you should use a URL starting with file://; e.g. alt.Chart(data="file://C:/Program Files/Python/Python36-32/lib/site-packages/vega_datasets/data/stocks.csv")

But since you said above that you were trying to pass the dataframe itself as data, I suspect that you somewhere mixed up your variable names. Make certain that what you are passing to the chart is actually what you think it is; passing a dataframe should work correctly whether it’s loaded from the web or from a local file.

[jakevdp]

The fig.to_dict() line is not doing anything here; it can be removed (it returns a dictionary representation of the chart without modifying the state of the chart).