宽幅 CSV 与 Plotly 快递

Question

I have a wide csv file that where the first column is the date (in Ymd format) and following columns, with headers are different covid19 related indicators, like in this example:

data	casos_novos_t	casos_novos_d	obitos_t	obitos_d
2021-21-04	123000	12000	34000	345
2021-22-04	134000	14000	34505	567

and so on. The sample dataset can be found here

I have the following code in Python:

# Import Libraries

import pandas as pd 
import plotly.express as px 

# Read CSV file 

df = pd.read_csv("covid19pt_data.csv")

# Plot

fig = px.bar(df)

fig.show()

After running it I get the following error message:

ValueError: Plotly Express cannot process wide-form data with columns of different type

If I change my code to

fig = px.bar(df, x='data',y='novos_casos_t')

fig.show()

the code works, showing me a bar graph for that column. The same for any other column.

However, after reading this post I was under the impression that plotly now supports wide format frames so I don't understand what I'm clearly doing wrong.

My experience level is close to zero and any help in making me understand what I'm doing wrong is much appreciated.

Answer 1

If you scroll down to the section on Wide-Form Defaults in Plotly , they explain that if you don't provide the parameters x or y for the px.bar method, then the default behavior is to set x to df.index and y to df.columns . However, your DataFrame is not indexed by date, and your date column data is the first column in df.columns

If we look at df.dtypes for your DataFrame:

data             object
casos_novos_t     int64
casos_novos_d     int64
obitos_d          int64
obitos_t          int64
d_internados      int64
d_uci             int64
recuperados_t     int64
recuperados_d     int64
vigilancia_t      int64
vigilancia_d      int64
ativos_t          int64
ativos_d          int64

We can see that the columns are not the same data type because the data column is an object type while the others columns are int64 , and since df.columns is passed to the y parameter by default, this is what is causing Plotly express to throw ValueError: Plotly Express cannot process wide-form data with columns of different type

If you want to use this default behavior to plot all of the numerical columns against the data column in a bar chart, you can set the index of your DataFrame to the data column using df = df.set_index('data') . This ensures that the remaining columns in your DataFrame are of the same int64 type.

# Import Libraries

import pandas as pd 
import plotly.express as px 

# Read CSV file 

df = pd.read_csv("covid19pt_data.csv")

## set the index as the data column
## all of the remaining columns are now of the same int64 type
df = df.set_index('data')

# Plot
fig = px.bar(df)

fig.show()