Issue when reprojecting multiple CSV files about dask-geopandas HOT 2 OPEN

# make some data
import os
import pandas as pd
import numpy as np

make_data=True
lat_lon_size=100000
csvs_num=10

if make_data:
    #first make some data
    np.random.seed(1234)
    #
    lat=np.random.random(lat_lon_size)*(10) - 37
    lon=np.random.random(lat_lon_size)*(10) - 64
    #
    df_=pd.DataFrame({'id':1,'Latitude':lat,'Longitude':lon})
    # dump dir
    os.makedirs('./csv',exist_ok=True)

    for i,f in enumerate(np.random.random(csvs_num)):
        df_.sample(frac=f).to_csv('./csv/test_{}.csv'.format(str(i)),index=False)


### Issue code 
import dask.dataframe as dd
import dask_geopandas
        
# from issue  just replace *_timeseries.csv' by test_*.csv
df = dd.read_csv('csv/test_*.csv')
gdf = dask_geopandas.from_dask_dataframe(df)
gdf = gdf.set_geometry(
    dask_geopandas.points_from_xy(gdf, x='Longitude', y='Latitude')
).set_crs('epsg:4326').to_crs('epsg:3395')
#
gdf.compute()