数据框列不会转换为 python 中的浮点数据类型

Question

I'm trying to get a percent change between two columns in a pandas dataframe that was built from an sqlite database, however when I try to do the calculations I cant because the columns are not of the float type. i have tried most ways of converting them like the .astype , to_numeric I've also tried to wrap it in float() , but none of this seems to work the data remains as a object type and I can do any computations on it.

conn = sqlite3.Connection("PriceData.db")
c = conn.cursor()
c.execute('SELECT productId, edition, lowPrice, midPrice, highPrice, date FROM PriceData WHERE date = "2020-07-30"')
data = c.fetchall()
columns = ['productId','edition', 'lowPrice','midPrice', 'highPrice', 'date']
dfPrev = pd.DataFrame(data, columns=columns, dtype = str)
conn = sqlite3.Connection("PriceData.db")
c = conn.cursor()
c.execute('SELECT productId, edition, lowPrice, midPrice, highPrice, date FROM PriceData WHERE date = "2020-07-31"')
data2 = c.fetchall()
dfToday = pd.DataFrame(data2, columns=columns, dtype = str)

result = pd.merge(dfPrev,dfToday, on = ['productId', 'edition'])

result['lowPercentChange'] = result[['lowPrice_x', 'lowPrice_y']].apply(lambda x :percentChange(result['lowPrice_x'], result['lowPrice_y']), axis = 1)
print(result)

print(result.dtypes)

Answer 1

have check the Type of data2. and it seems like you are passing str as dtype argument in pd.Dataframe, have you tried to change into float? Please let me see the picture exactly and your output result, might be I am not understanding correctly.