marvande / epfl-ada-2019-data-analysis Goto Github PK

maybe doxygen ?

Cleaning the People dataset, making sense of household size, number of kids, marital status and household composition.

Have a look and think about some simple plots we can do to have a better idea of what could be some trends or interesting things in the data.

We focused until now on understanding the relationship between the demographic data and the transactions of the consumers. Thus, we had to filter out a lot of data because demographic data was only available for approximately one third of the households. Now, we saw that there are some correlations between product quantities, and we would like to know if they are accentuated in the whole transaction data.

Choose from the user's guide which datasets we want. Write a short readme describing each dataset (column names, content, etc).

Assign numbers or letters to different categories so that we can plot them and sort them more easily.

Clean up the transaction dataset on different jupyter notebook. Try to find budgets and total amount spent per household.

Load datasets into panda dataframes. Add names of dataframes to df_names.md

Write a tutorial for ZenHub and add it to the doc folder.

Run again the Products notebook or fix it