MegatenFreak Active Member Licensed User Apr 7, 2021 #1 Hi. I do some additions in my app, and I ran into this strange problem: when I add 1.3 and 0.1, I get 1.4000000000000001 instead of just 1.4: Code: Copy Log(1.3 + 0.1) Log(1.3 + 0.11) Log(1.3 + 0.102) All of the above result in a large number just like the example. What is strange is that IT ONLY HAPPENS WHEN "3" IS INVOLVED. All other numbers (1.4, 1.5, etc.) work just fine. I even defined the numbers separately as doubles and then added them. same result. I even tried them as strings. Please help... I'm so confused.
Hi. I do some additions in my app, and I ran into this strange problem: when I add 1.3 and 0.1, I get 1.4000000000000001 instead of just 1.4: Code: Copy Log(1.3 + 0.1) Log(1.3 + 0.11) Log(1.3 + 0.102) All of the above result in a large number just like the example. What is strange is that IT ONLY HAPPENS WHEN "3" IS INVOLVED. All other numbers (1.4, 1.5, etc.) work just fine. I even defined the numbers separately as doubles and then added them. same result. I even tried them as strings. Please help... I'm so confused.
Erel B4X founder Staff member Licensed User Longtime User Apr 7, 2021 #2 The binary representation cannot 100% accurately represent all decimal numbers. Never convert a number to a string without using NumberFormat (or one of the similar features). Upvote 0
The binary representation cannot 100% accurately represent all decimal numbers. Never convert a number to a string without using NumberFormat (or one of the similar features).
J jerry07 Member Licensed User Longtime User Apr 7, 2021 #3 If you need to show this as 1.4 and not 1.40000000001 here is example. Not sure if there any other better ways to address this. B4X: Copy value1 = 1.3 + 0.1 Log(NumberFormat2(value1, 1, 2, 0, False)) The 3rd parameter will specify number of decimals to show. Upvote 0
If you need to show this as 1.4 and not 1.40000000001 here is example. Not sure if there any other better ways to address this. B4X: Copy value1 = 1.3 + 0.1 Log(NumberFormat2(value1, 1, 2, 0, False)) The 3rd parameter will specify number of decimals to show.
MegatenFreak Active Member Licensed User Apr 7, 2021 #4 I use NumberFormat (and NumberFormat2) a lot. It's a life saver! I was just wondering if there's a particular reason, something evil or mysterious about number 3, or if I was doing something wrong. Fortunately, thanks to Erel, NumberFormat also "rounds" the number to the specified number of fractions, so it also fixes the problem I had with such results as 1.399999... (p.s. Just in case anyone is interested, this problem arose when I was writing a function to do a "banker" rounding (round to even) on numbers. This method works differently when it tries to round a 5: it rounds it to the nearest higher even number, so, for example, 0.65 rounds to 0.6, but 0.75 rounds to 0.8) Upvote 0
I use NumberFormat (and NumberFormat2) a lot. It's a life saver! I was just wondering if there's a particular reason, something evil or mysterious about number 3, or if I was doing something wrong. Fortunately, thanks to Erel, NumberFormat also "rounds" the number to the specified number of fractions, so it also fixes the problem I had with such results as 1.399999... (p.s. Just in case anyone is interested, this problem arose when I was writing a function to do a "banker" rounding (round to even) on numbers. This method works differently when it tries to round a 5: it rounds it to the nearest higher even number, so, for example, 0.65 rounds to 0.6, but 0.75 rounds to 0.8)
agraham Expert Licensed User Longtime User Apr 8, 2021 #5 If you need decimal precision BigNumbers library | B4X Programming Forum Upvote 0
MegatenFreak Active Member Licensed User Apr 8, 2021 #6 That is awesome! I wish I had done some research before banging my head against the wall for hours!!! Upvote 0
That is awesome! I wish I had done some research before banging my head against the wall for hours!!!
