The fact that tocol of the banana map already gives you things in the preferred order is a game changer, I made a huge let with sortby for absolute distance, then rows and columns when I really didn't need to! Very clever solution, I'm going to try and find some other solves for that one to see if there are other methods but your one was really good there!
Awesome idea for managing the mine! Still need to properly watch your structure for the cards, but I start suspecting one of the following: a) there is probably a significant shortcut to solve it or b) it was easier to do manually
I think it would be pretty hard to do manually… 52 cards to go through, and your hand starts getting pretty big after a while! Then again, with some semi-automation, it might still not be the worst approach…
If I'm reading the results correctly, it seems only one participant (Cédric Morgan Rasolofo) scored any points at all on it during the live run! And not only that, he got full points - would be very interesting to see what he did.
Here is my level 5 recursive lambda: Level 5: I did use a recursive lambda: input is an array that consists of lana's current cell vstacked with the addresses of all the bananas. =LAMBDA(input, LET(bananas, DROP(input, 1), lana, INDEX(input, 1), banDist, BYROW(bananas, LAMBDA(a, _distance(a, lana, MAX))), "- Calculates distances to all bananas borderDist, MIN(_rowCol(lana) - 1, 41 - _rowCol(lana)), "- Distance to the border nearBan, MIN(banDist), "- Don't need this based on your video. ind, XMATCH(MIN(banDist), banDist), before, TAKE(bananas, ind - 1), "- All bananas before the nearest after, TAKE(bananas, ind - ROWS(bananas)), "- All bananas after the nearest newBananas, IF(ind = 1, after, IF(ind = ROWS(bananas), before, VSTACK(before, after))), target, INDEX(bananas, ind), newInput, VSTACK(target, newBananas), ans, IF(borderDist < nearBan, 0, 1 + _lana(newInput)), ans))
Bonus 1, if I was solving it in competition, I would NOT have added the markup, as the wording of the question implied to me that it wasn't supposed to include a markup, it asked for the cost to produce, whereas in lvl 1 it asked for markup value. Level 2, my first read I thought ALL mining cost energy, not just the uranium and mercury. Again, skim reading is an issue! My formula for lvl 2: =LET(stRow, AQ19, poss, TAKE(DROP(MINE, stRow - 2), 4), seq, HSTACK(CHOOSEROWS(poss, 1), IFERROR(reverse(CHOOSEROWS(poss, 2)), " "), IFERROR(CHOOSEROWS(poss, 3), " "), IFERROR(reverse(CHOOSEROWS(poss, 4)), " ")), mined, FILTER(seq, seq " "), toxic, SCAN(0, mined, LAMBDA(a,v, IF(OR(v = "☢", v = "🌡"), a + 1, a))), life, 100 - SCAN(0, toxic, SUM), spoils, TAKE(mined, 1, SUM(--(life > 0)) + 1), coins, XLOOKUP(spoils, minerals, values), SUM(coins)) For seq, I assumed that you would never need more than 4 rows. Level 3 messed me up by having the 25 as a string instead of a number. But, I wasn't on a timer, so I had time to fix it.
I agree, I thought B1 wasn’t super clear that way (I even asked about it on the live stream). That’s a nice solution to L2! Looks like our approaches were pretty similar, except that I didn’t do the snake as part of the single cell.
The fact that tocol of the banana map already gives you things in the preferred order is a game changer, I made a huge let with sortby for absolute distance, then rows and columns when I really didn't need to! Very clever solution, I'm going to try and find some other solves for that one to see if there are other methods but your one was really good there!
I’m looking forward to seeing what you came up with - especially the simple solution to the cards!
@@DimEarly Just uploading now, I basically searched if the possible sequences were in the string then took them out
great content, thx
I like your technique of FIND(LEFT( ), "SDT") that used to convert the Dart record to score. I will try that out. Thanks!
Totally stolen from Bo (@ExcelWizard) - but it was a very smart trick!
Awesome idea for managing the mine!
Still need to properly watch your structure for the cards, but I start suspecting one of the following: a) there is probably a significant shortcut to solve it or b) it was easier to do manually
Personally I setted a list of possible combinations manually
I think it would be pretty hard to do manually… 52 cards to go through, and your hand starts getting pretty big after a while!
Then again, with some semi-automation, it might still not be the worst approach…
12:27 This case was no way easy. 😅 Thank you for sharing!
Amen to that! 😂
1:11:55 The worst part is that I feel the same energy at that moment😂😭
That card case was no joke
Amen! Although I do think my structure made it harder than it needed to be…
If I'm reading the results correctly, it seems only one participant (Cédric Morgan Rasolofo) scored any points at all on it during the live run! And not only that, he got full points - would be very interesting to see what he did.
*Hello and welcome!*
🏆💻📊📉 😎
Keep going
Here is my level 5 recursive lambda:
Level 5:
I did use a recursive lambda: input is an array that consists of lana's current cell vstacked with the addresses of all the bananas.
=LAMBDA(input,
LET(bananas, DROP(input, 1), lana, INDEX(input, 1),
banDist, BYROW(bananas, LAMBDA(a, _distance(a, lana, MAX))), "- Calculates distances to all bananas
borderDist, MIN(_rowCol(lana) - 1, 41 - _rowCol(lana)), "- Distance to the border
nearBan, MIN(banDist), "- Don't need this based on your video.
ind, XMATCH(MIN(banDist), banDist),
before, TAKE(bananas, ind - 1), "- All bananas before the nearest
after, TAKE(bananas, ind - ROWS(bananas)), "- All bananas after the nearest
newBananas, IF(ind = 1, after, IF(ind = ROWS(bananas), before, VSTACK(before, after))),
target, INDEX(bananas, ind),
newInput, VSTACK(target, newBananas),
ans, IF(borderDist < nearBan, 0, 1 + _lana(newInput)), ans))
Bonus 1, if I was solving it in competition, I would NOT have added the markup, as the wording of the question implied to me that it wasn't supposed to include a markup, it asked for the cost to produce, whereas in lvl 1 it asked for markup value.
Level 2, my first read I thought ALL mining cost energy, not just the uranium and mercury. Again, skim reading is an issue!
My formula for lvl 2:
=LET(stRow, AQ19,
poss, TAKE(DROP(MINE, stRow - 2), 4),
seq, HSTACK(CHOOSEROWS(poss, 1), IFERROR(reverse(CHOOSEROWS(poss, 2)), " "), IFERROR(CHOOSEROWS(poss, 3), " "), IFERROR(reverse(CHOOSEROWS(poss, 4)), " ")),
mined, FILTER(seq, seq " "),
toxic, SCAN(0, mined, LAMBDA(a,v, IF(OR(v = "☢", v = "🌡"), a + 1, a))),
life, 100 - SCAN(0, toxic, SUM),
spoils, TAKE(mined, 1, SUM(--(life > 0)) + 1),
coins, XLOOKUP(spoils, minerals, values),
SUM(coins))
For seq, I assumed that you would never need more than 4 rows.
Level 3 messed me up by having the 25 as a string instead of a number. But, I wasn't on a timer, so I had time to fix it.
I agree, I thought B1 wasn’t super clear that way (I even asked about it on the live stream).
That’s a nice solution to L2! Looks like our approaches were pretty similar, except that I didn’t do the snake as part of the single cell.