Upload and merge multiple CSV files to one DataFrame

Using Streamlit
1

Hey there, great job on now having the accept_multiple_files flag for file_uploader. However, I am having an issue when I want to upload multiple CSV files, and then concatenate these files into one DataFrame.

uploaded_files = st.file_uploader("Upload CSV", type="csv", accept_multiple_files=True)
if uploaded_files:
    uploaded_data_read = [pd.read_csv(file) for file in uploaded_files]
    raw_data = pd.concat(uploaded_data_read)

returns

EmptyDataError: No columns to parse from file

for the line [pd.read_csv(file) for file in uploaded_files].

I don’t get issues when I upload one file, only when I upload more than one file.

Here is the output of st.write(uploaded_files) after uploading two files

[
  "<class 'streamlit.uploaded_file_manager.UploadedFile'>",
  "<class 'streamlit.uploaded_file_manager.UploadedFile'>"
]

Can anyone help?

2

Try:

uploaded_files = st.file_uploader("Upload CSV", type="csv", accept_multiple_files=True)
if uploaded_files:
    for file in uploaded_files:
        file.seek(0)
    uploaded_data_read = [pd.read_csv(file) for file in uploaded_files]
    raw_data = pd.concat(uploaded_data_read)

The UploadedFile class is a subclass of BytesIO, and therefore it is “file-like”. This means you can pass them anywhere where a file is expected. As a subclass of BytesIO, make sure to reset the buffer after reading it with UploadedFile.seek(0).

2 Likes
3

@nigani

Is it possible to create a similar approach that allows to upload multiple txt files and concatenate them into one, and pass down for subsequent steps. Is it possible to write the concatenated file in the tmp directory such that it get removed after being used by the web app.